Use of Evidence to Inform Agricultural Policy Decisions: What have we learned from experience in Africa? Authors: Christopher Delgado, Karen Brooks, Christian Derlagen, Steven Haggblade, and Kate Lawyer i ii Use of Evidence to Inform Agricultural Policy Decisions: What have we learned from experience in Africa? Authors: Christopher Delgado, Karen Brooks, Christian Derlagen, Steven Haggblade, and Kate Lawyer iii © 2019 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover illustration and design by Brad Amburn Acknowledgements This paper was researched and written by a team consisting of Christopher Delgado (Consultant, World Bank Africa Agriculture Policy Unit), Karen Brooks (Independent Consultant), Christian Derlagen (Independent Consultant), Steven Haggblade (Professor, Michigan State University), and Kate Lawyer (former Program Officer, Bill and Melinda Gates Foundation). Financial support from the Bill and Melinda Gates Foundation under grant ID OPP1208822 is gratefully acknowledged. Sincere thanks are also due to Holger Kray of the World Bank Group, under whose leadership this work was launched and carried out. This paper is a follow-up to the consultation meeting on Strengthening African Agriculture Policy Capacity that was hosted by the World Bank Group (WBG) Agriculture Global Practice and the Bill and Melinda Gates Foundation (BMGF) in Washington, D.C. on May 1-2, 2019. The consultation meeting was part of a collaboration between the WBG, BMGF, and the Alliance for a Green Revolution in Africa (AGRA) on the issue of building capacity to better use evidence in agricultural policymaking in Africa. For convenience, the executive summary of the minutes of the workshop is reproduced in Annex I of the present paper. The full workshop report is referenced under Sakane, Delgado, and Rege (2019) in the list of references and is available on demand. The authors extend sincere thanks to the many people who contributed to the paper as it was written after the workshop. These contributors include David Atwood and colleagues at United States Agency for International Development (USAID) Feed the Future; Ousmane Badiane, Tsitsi Makombe, and Kimsey Savadogo of the International Food Policy Research Institute (IFPRI); Simeon Ehui and Holger Kray of the WBG; Thom Jayne of Michigan State University; Kate Lawyer, Joshua Ariga, and colleagues at BMGF; Ferdi Meyer of the University of Pretoria and Bureau for Food and Agricultural Policy; and Jonathan Said of the Tony Blair Institute. Beyond the helpful comments on parts or all of the manuscript, some of these colleagues contributed paragraphs to the narrative. While many of their comments were included, these colleagues did not specifically endorse the paper and may or may not agree with some parts. The authors retain full responsibility for content. 1 Acronyms ACC Agricultural Commercialization Clusters Initiative ACF Agricultural Consultative Forum AFD Agence Française de Développement AfDB African Development Bank AGRA Alliance for a Green Revolution in Africa AGRODEP African Growth and Development Policy Modeling Consortium APME Agricultural Policy, Monitoring, and Evaluation Units APRU Agricultural Policy Research Unit ASMP Agricultural Sector Management Project ATA Agricultural Transformation Agency ATOR Annual Trends and Outlook Report AU African Union AUDA-NEPAD African Union Development Agency BFAP Bureau for Food and Agricultural Policy BHN Basic Human Needs BMGF Bill and Melinda Gates Foundation BR Biennial Review CAADP Comprehensive Africa Agricultural Development Programme CARD Centre for Agriculture Research and Development CCARDESA Center for Coordination of Agricultural Research and Development of Southern Africa CEPPAG Research Center for Agricultural and Food Policies and Programs CGIAR Consultative Group on International Agricultural Research COMES Common Market for Eastern and Southern Africa CSA Commissariat à la Sécurité Alimentaire CSSP Country Strategy Support Program DAFF Department of Agriculture, Forestry and Fisheries DfID Department for International Development DRC Democratic Republic of Congo EC European Commission Ecofil Economie des Filières ECOWAS Economic Community of West African States EPRC Economic Policy Research Centre EU European Union FANRPAN Food and Natural Resources Policy Analysis Network FAO United Nations Food and Agriculture Organization FAPRI Food and Agricultural Policy Research Institute FDI Foreign Direct Investment 2 FIRST Food and Nutrition Security Impact, Sustainability and Transformation FSRP Food Security Research Project GDP Gross Domestic Product HIPC Heavily Indebted Poor Countries IA Impact Assessment IAPRI Indaba Agricultural Policy Research Institute IER Institut d’Economie Rurale IFPRI International Food Policy Research Institute IFRA Institut de Formation et de Recherche Appliquée ILO International Labor Organization IPR Institut Polytechnique Rural IRES Institute of Social and Economic Research JSR Joint Sector Review KIPPRA Kenya Institute for Public Policy Research and Analysis M&E Monitoring and Evaluation MAFAP Monitoring and Analyzing Food and Agricultural Policies MARE Ministry of Agriculture and Animal Resources MaSSP Malawi Strategy Support Program MoA Ministry of Agriculture MoALR Ministry of Agriculture and Livestock Resources MoEF Ministry of Economics and Finance MSU Michigan State University NAIP National Agricultural Investment Plan NAMC National Agricultural Marketing Council OAU Organization of African Unity OMA Observatoire du Marché Agricole PAPA Projet d’Appui aux Politiques Agricoles PIU Project Implementation Unit ReNAPRI Regional Network of Agricultural Policy Research Institutes ReSAKSS Regional Strategic Analysis and Knowledge Support System SADC Southern African Development Community SAKSS Strategic Analysis and Knowledge Support Systems SAL Structural Adjustment Loan/Credit SAP Système d’Alerte Précoce TBI Tony Blair Institute for Global Change USAID United States Agency for International Development WBG World Bank Group 3 Table of Contents Foreword 8 Executive Summary 10 The challenge and why it matters 11 The changing historical context and needs that have arisen for policy evidence 11 Overview of the cases for building agricultural policy analysis and delivery systems examined in the paper 12 Take-aways from the analysis of cases since 1980 16 In conclusion 19 1. Challenge: The Weak Link from Increasing Available Evidence to Increased Use of It 20 2. Why Is Relevant Evidence More Important for Policymakers Now Than in the Past? What Has Changed? 24 2.1: The rising share of agricultural public expenditure from domestic sources 25 2.2: The rising need for accountability in an era of tightening domestic budgets 25 2.3: Given increased need, why is evidence underused in agricultural policymaking? 28 3. The Evolving Context: 1960 to 1990 30 3.1: Addressing the educational preconditions 31 3.2: The transformative 1970s: the oil shock of 1973, food crisis of 1974, and expansion of foreign assistance to African agriculture 33 3.3: “Structural Adjustment” in the 1980s and 1990s 35 4. Capacity-Building for Agricultural Policy Analysis and Advice Since 1980 38 4.1: The ascent of country-specific agricultural policy analysis units after 1980 39 4.2: Seeking external advice on agricultural policy without building new institutions 42 4.3: Regional approaches to improving the use of evidence in agricultural policy 43 4 5. Case studies 46 5.1: Agricultural policy and monitoring units within MoA in the 1980s and 1990s: the Tanzania Agricultural Sector Management Project 47 5.2: Agricultural policy think-tanks or agencies derived from external university-to-domestic university programs in Kenya, Zambia, and Mali 49 5.3: An independent agricultural think tank linking universities, the private sector and government: The South African Bureau for Food and Agricultural Policy 54 5.4: Externally-led extended term agricultural policy analysis units in-country: the country programs of IFPRI 57 5.5: Network strengthening approaches for MoAs where domestic analysis capacity is widespread: the PAPA Project in Senegal’s MoA 61 5.6: On-call technical assistance from multilaterals having centralized funding: the Monitoring and Analyzing Food and Agricultural Policies (MAFAP/FAO) 63 5.7: High-level embedded foreign agricultural policy advisors in MoAs: the EU/FAO FIRST Policy Assistance Facility 66 5.8: High-level embedded foreign agricultural policy advisors in the Executive: the Tony Blair Institute for Global Change (TBI) 68 5.9: Government agricultural policy and monitoring units outside MoA: the Ethiopian Agricultural Transformation Agency 70 5.10: An African-led regional policy analysis network to strengthen CAADP: ReSAKSS 72 5.11: An African-led regional policy analysis network built around a university identity: ReNAPRI 76 6. Synthesis of Lesson-Learning from the Cases 80 6.1: Ownership by national decision-makers and trust matter 82 6.2: Longevity matters in the establishment of trust 82 6.3: The record of delivery units within a Ministry of Agriculture has not been impressive 84 6.4: Implementation of agricultural strategy based on evidence requires the ability to work with entities beyond MoAs 85 5 Table of Contents 6.5: Placement within a Prime Minister’s or President’s office can undermine sustainability 86 6.6: Independent think tanks are good for policy research, but face challenges in moving research findings into use in the policy process 86 6.7: Embedding foreign advisors is preferable to short-term technical assistance missions but should include specific efforts to build national capacity 88 6.8: Regional networks can help build ownership in defining regional policy options while capturing economies of scale and scope from a regional approach 88 6.9: Success should be measured by contribution and influence rather than policy outcomes 89 6.10: The costs of supporting trusted agricultural policy analysis are modest 89 7. Conclusions 92 References 95 Annex I 100 Consultation Meeting on Strengthening African Agriculture Policy Capacity 101 Lessons learned to date 102 ‘Principles’ for institutional architecture going forward 104 Endnotes 107 List of Tables Table 1: Share of Agriculture Public Expenditures Financed by Donors 26 List of Figures Figure 1: Government debt as a share (%) of GDP for African countries, 2017 27 Figure 2: BFAP policy engagements over 15 years and policy outcomes 56 6 Foreword African governments and inter-governmental organizations have made increasingly clear the collective determination on the continent for improving the welfare of all citizens and self-determination in how to go about it. There has also been strong recognition of the importance in this regard of African ownership of reliable and adequate data, requisite analytical skills, and strong institutions for creating and curating knowledge. These are required for elucidating alternative policy options and their likely consequences, both positive and negative. Nowhere has this been more evident than in the case of agriculture and food security, as witnessed since 2003 by strong national and regional ownership of Africa’s own Comprehensive Africa Agriculture Development Programme (CAADP) and associated activities. And tremendous progress has been made in recent decades to improve the generation of evidence-based agricultural policy analysis in Africa through investments in higher education, policy research capacity, and better data. Yet a recurring refrain we encounter from agricultural policymakers is that at the time they need to make decisions, they do not have access to impartial analysis dealing credibly with the costs and benefits of alternative actions in a form they can use at the time needed. Indeed, building the capacity of national governments to use existing evidence-based policy analysis has been sadly neglected in recent years. Fortunately, this neglect is now what many governments increasingly seem to want to change, even as the share of agricultural and food investments—both 8 public and private—funded from domestic sources increases. And the stakes for Africa of getting evidence-based policy support functions right in agriculture have gone up in the current context of conflict, fragility, climate change, feeding growing cities, and the critical need for more and better jobs for youth on a scale that requires major contributions from growing agricultural value chains. Lessons of experience summarized in this report show that the process for addressing these policy support gaps matters nearly as much as the immediate results obtained; ownership by those concerned is key. We believe that many African governments and their partners wish to invest in a solid national process to use existing domestic policy data and analysis capacity better to improve agricultural outcomes and make them more forward-looking. The pages that follow suggest opportunities and pitfalls to be aware of in embarking on such a course. Simeon K. Ehui Regional Director for Africa Sustainable Development Practice Group The World Bank 9 ES 10 Executive Summary2 The challenge and why it matters Agricultural policymakers in Africa increasingly face the need for policy options based on evidence-based analysis to promote agricultural transformation and to adapt to climate change. Furthermore, data and analytical tools to support informed agricultural policymaking are increasingly abundant thanks to investment in these areas, mostly from external sources. Still, the use of hard data and robust analyses linked to outcomes (“evidence”) are still rare in most agricultural policymaking in the region. The present paper focuses on what can be learned to improve outcomes from experiences promoting the increased use of evidence in agricultural policymaking. Today, ministries of agriculture (MoAs) are increasingly under pressure to show ministries of economy and finance (MoEFs) both the rationale behind proposed spending and the impact of past spending, particularly net estimated impacts on forex and fiscal balances. Even so, at present most African governments are still under-spending on agricultural public goods such as research, extension, and infrastructure. This is relative to other parts of the developing world on both a share basis and an absolute basis per capita. Many countries still overspend on short-term agricultural subsidies to individuals. The balance between capital and recurrent expenditures is often skewed, with diminished returns to each. Further, agricultural budgets are often under- spent at the end of the fiscal year, jeopardizing new allocations to agriculture. Addressing these problems will require proactive institutional change in how policy is made and implemented, including more use of empirical evidence in policymaking and the monitoring of implementation. The changing historical context and needs that have arisen for policy evidence The availability of large-scale foreign funding for higher education during the Cold War permitted the launch of a series of African agricultural universities and research institutes across the continent. Optimism about agriculture and overall economic development prevailed in the 1960s; the macroeconomic 11 situation gave little cause for concern. But everything changed during the 1970s. The global and Sahel food crises and both a quick ramping up of foreign assistance and its fragmentation into many projects under many different sources of external decision-making caused turmoil. Funding growth and complexity outstripped the nations’ abilities to coordinate projects dispersed across rural areas. National governments also had to provide counterpart staff to each donor, which stretched demands on the scarce professional expertise of ministries. Concurrent with these developments, a devastating Africa-wide fiscal and balance-of-payments crisis came to a head in 1979. An urgent need for external financial assistance was met with demands for policy reforms and transparency, especially in agriculture. The latter became the center of macroeconomic policy and political discussions in a way it had not been before. Beginning in the early 1980s, governments—and donors at the behest of governments—funded efforts to promote the generation of evidence for agricultural policymaking and supported institutions that generated and housed that evidence. The efforts took various forms and were influenced not only by national differences and needs, but also by contextual factors such as the state of the national agricultural education and research systems. Overtime, a wide variety of institutional approaches for the generation of core evidence relevant to agricultural policy have been tried without comparable emphasis on final delivery of the products decision-makers needed. Considerable progress has been achieved in growing the capacity for collecting relevant data, analyzing it, and producing first class and in-depth agricultural policy research in Africa since 1980. Yet, the delivery of this improved evidence to policymakers needs to include very succinct, standardized assessments of costs, benefits, and likely spillover effects of alternative options. This information needs to get to policymakers on time and to be trusted by them as accurate and impartial. The present paper focuses on experience with these final delivery mechanisms for evidence in Africa and lessons that can be drawn from them. Overview of the cases for building agricultural policy analysis and delivery systems examined in the paper Several MoEFs encouraged donors to fund agricultural policy, monitoring, and evaluation (APME) units directly in MoAs in the 1980s and 1990s. This form of institutional arrangement responded to the view at the time that public 12 entities could and should occupy the center of the policymaking chain and address issues of implementation. These AMPE unites provide institutional buy-in and institutional memory within governments. This paper discusses a well-funded and well-evaluated case in Tanzania. Agricultural policy analysis think tanks or policy research institutes were established in the 1980s and 1990s in affiliation with universities in countries with strong, existing agricultural institutions. These functioned both to generate policy evidence as well as to deliver recommendations to policymakers. This paper reviews the Agricultural Policy Research Unit (APRU) at Bunda College in Malawi and the Tegemeo Institute at Edgerton University in Kenya. One of the benefits of these institutions was the freedom to pay higher salaries and thus to retain more qualified staff. A key element to success was the freedom to tap into funding from a variety of donors, including some donors who had very specific asks, but who were not willing to contribute beyond the variable marginal costs of the output they wanted. In the 1990s, donors, including the World Bank, moved to support national efforts to establish macroeconomic policy think-tanks serving the needs of MoEFs in the analyses of structural adjustment policies, which included particular attention to agricultural incentives. These think tanks were largely separate from line ministries and had special relationships with government institutions. Examples discussed include the Economic Policy Research Centre (EPRC) in Uganda, founded in 1993, and the Kenya Institute for Public Policy Research and Analysis (KIPPRA), founded in May 1997. The lesson for the current paper is that both of these institutions survived and remain involved in informing policy choices under national MoEF sponsorship, although their attention to agricultural issues is much diminished. This contrasts with the experience since the 1990s of agricultural units under MoAs and university think tanks. A country-level model for channeling evidence from domestic think tanks and NGOs into agricultural policy has surfaced recently in Senegal. Senegal has nine public and nongovernmental institutions carrying out various forms of agriculturally-relevant policy research and analysis, little of which was being used systematically for policymaking by the MoA prior to the launch of the Projet d’Appui aux Politiques Agricoles (PAPA) project in 2016. PAPA, housed in the MoA, has a mandate to communicate MoA policy priorities and specific queries to the network of nine domestic policy analysis institutions, to 13 facilitate their access to funding, and to curate their output for policymakers. The model has been successful in this regard. However, it is difficult to assess if PAPA can continue in its present form without substantial external funding and skilled analytical and organizational input by external agencies not bound by domestic civil service rules. In 2003, in response to coordination issues in funding, IFPRI commenced a series of strong Country Strategy Support Programs (CSSP) funded by multiple projects and donors across a variety of topic areas in each country, which where run as a coordinated whole. These country programs had varying degrees of closeness to host country governments. This was a case of coordination of donors by the agency being funded and is explored in detail in this paper. “Implementation Units” are a feature of the projectization of foreign assistance. In agriculture they are typically embedded within an MoA. As the name suggests, project implementation units (PIUs) focus on implementation and delivery of project results based on previous decisions. They are well-placed to feed evidence into ongoing decision-making on policy. Where the number of projects is large, coordination needs to be provided by a similar, over-arching unit that answers directly to government. The solution adopted in Ethiopia was an independent government agency, the Agricultural Transformation Agency (ATA), outside the Ministry of Agriculture. ATA can hire on a merit basis, pay higher salaries, and better retain key staff. ATA is a solution best adapted to a large country with a large agricultural investment program, a high degree of external support to agricultural public investment, a large supply of high caliber talent in economics and agricultural policy, and tolerance for a high degree of government intervention in agriculture. As will be discussed, following such a model requires considering the sustainability of the approach, whether it should be only a temporary solution, and the unintended consequences for the performance of other government agencies. Countries may want technical input in a specific area such as an agricultural public expenditure review, but not want to support the capacity for such work nationally. The United Nations Food and Agriculture Organization (FAO) provides short-term, on-call technical assistance in agricultural policy analysis to individual countries that request it. Core funding for such work has become increasingly problematic. However, the rise of substantial trust funding has extended the feasibility of the model. The Monitoring and Analyzing Food 14 and Agricultural Policies (MAFAP) program—reviewed in this paper—is trust funded by donors since 2009 and housed within FAO. It provides in-depth assessment of agricultural policy issues on the basis of short-term missions to the countries concerned. As an alternative to task-specific, short-term technical assistance, embedded foreign technical advisors—common shortly after independence in many countries in the 1960s—are again being placed within MoAs in Africa. Embedded foreign advisors that cover food and agriculture typically deal with recommendations on policy and strategic issues at a high level. Bilateral donors such as the United Kingdom’s Department for International Development (DfID), the Agence Française de Développement (AFD), and the United States Agency for International Development (USAID) have such personnel in selected countries. The European Commission (EC) has a major trust funded program allowing the placement of senior technical-level FAO advisors in 12 African countries through the Food and Nutrition Security Impact, Sustainability, and Transformation Policy Assistance Facility (FIRST), discussed in the paper. The Tony Blair Institute (TBI) for Global Change, also discussed in the paper, places foreign advisors at high levels of government with teams often anchored in the Office of the President. Unlike the bilateral donor and FAO models for embedded advice that typically focus on policy, the TBI model focuses primarily on management, coordination, and delivery. Mobilizing regional resources to support national agricultural programs is an approach to improving the use of evidenced-based analysis and building African ownership. The Comprehensive Africa Agriculture Development Program (CAADP) process was launched in Maputo in 2003 and was confirmed by the African Union. It came about in the context of calls for more strategic, more evidence-based, and more inclusive designs for public agricultural investments. The Regional Strategic Analysis and Knowledge Support System (ReSAKSS), discussed in this paper, supports CAADP. It consolidates and curates data relevant to agricultural strategy and promotes a network of African agricultural policy researchers that contribute to analysis and peer review. In this sense, CAADP and the ReSAKSS initiative contribute both to the generation of evidence and to increasing its actual use by African MoAs. Finally, in policy analysis and advice, as in agricultural research more generally, most African countries suffer from a “small country” problem. 15 Where needs exceed national capacity and are replicated across borders, regional networks and alliances can leverage competence and enhance results. The Regional Network of Agricultural Policy Research Institutes (ReNAPRI), discussed in this paper, is a coordinated group of national agricultural policy research institutes in Southern and Eastern Africa that undertakes agricultural policy research and analysis. Activities associated with ReSAKSS and ReNAPRI facilitate access of governments to proven African skills in this area and foster positive spillovers across countries. Take-aways from the analysis of cases since 1980 Ownership by national decision-makers and trust matter. Stakeholders in the policy process must be confident that they can make their needs known, and that policy analysts will address these needs with professionalism and timeliness. In this sense, the effort must be owned by stakeholders, particularly those with the greatest power in the policy process. This point is emphasized time and again in the works and consultations summarized in this paper. Given the importance of ownership, it is ominous that so few of the examples reviewed were or are now funded nationally. Most depend substantially or wholly on external funding. Finding a solution to this financial challenge is essential to finding a solution to the problem at hand. Longevity matters in the establishment of trust. Since policymakers will not in general be well-placed to evaluate the methodology leading to results, they must have trust in the researchers and particularly in the management of the underlying research effort from the standpoints of quality and objectivity. The cases illustrate the long time required for the establishment of a solid reputation that underpins the trust of decision-makers. Longevity matters particularly if the intent is to have a steady flow of evidence to inform policymaking over an extended time horizon. If the intent is instead to seek advice on a specific and time-bound issue, then external technical assistance can also be commissioned. The long-term performance of policy analysis units within MoAs has not been impressive. Efforts to create policy support functions of the type discussed here as units within African MoAs have foundered for several reasons. Civil service pay-scales and rules that govern MoAs do not in general serve to recruit and retain staff with highly specialized skills in demand outside of government. Governments and donors facing these constraints have sometimes resorted 16 to moving defined functions to parastatals to allow different pay scales and promotion opportunities. The organizational culture within MoAs also often stresses politics, process, and practice, rather than promoting the technical and analytical excellence required for good policy analysis. Implementation of agricultural strategy based on evidence requires working with entities beyond MoAs. An agency feeding into agricultural policy decisions needs to have strong links to different ministries, parliamentary committees, media, and trade associations. It also needs to be in touch with the realities of implementation. The ATA in Ethiopia has such a capacity to work across governmental units or with outside actors. Experience presented in the cases suggests that having such an institution in government can create frictions with line ministries. If an enclave entity is considered a short-term measure to cover acute needs while capacity within the line ministries is enhanced, then the plans for capacity development become important, as does the exit strategy. The sustainability of such institutions can be an issue if they are reliant on external funding. Direct placement within a Prime Minister’s or President’s office of a unit to improve coordination can undermine sustainability. Creation of an agricultural policy support function directly under the chief executive, but without adding a large new institution such as ATA, can facilitate coordination between and among line ministries and can also assure high-level access to policymakers. However, such placement close to a locus of political power while still relying on line ministries for oversight of implementation can lead to disconnects between policymaking and policy implementation. It can also raise perceptions of politicization in agricultural policy, thereby undermining trust more widely within the policy ecosystem. Moreover, close affiliation with the head of an administration can undermine longevity by reducing the likelihood that the unit will weather a change in administrations. Independent think tanks are good for policy research, but face challenges in moving research findings into use in the policy process. Agricultural policy research think-tanks are often strong models for identifying policy research issues that will be important down the road, for fundraising, for motivating key national talent, and for generating evidence that is free of political bias. However, these think tanks are typically not well structured to address the weak link between evidence generation and its use as defined in this paper. 17 Analytical work from think tanks is typically designed without explicit reference to policy decisions affecting specific actors that would occur according to a known calendar. It thus often misses key entry points for serving decision- makers. Further, think tanks reliant on external funding from diverse sources are vulnerable to fragmentation of their priorities and erosion of institutional coherence, a problem that goes beyond Africa. The roles of generating in-depth research results, foresight, and preparing for future policy needs should be understood as necessary components of a national agriculture policy ecosystem. However, they are not sufficient by themselves for the use of evidence in agricultural policymaking as they do not have a good track record of getting evidence in front of policymakers in the right form and at the right time. Embedding foreign advisors is preferable to short-term technical assistance missions, but should include specific efforts to build national capacity. Using externally funded and sourced embedded advisors in national agricultural policy processes can be useful for short-term assistance to addressing institutional weaknesses. Embedded advisors are generally preferable to short-term externally sourced and funded technical assistance missions. However, they are most useful when combined with a specific institutional capacity-building effort to permit nationalization of the embedded advisors’ functions in the long term. Regional networks can help build ownership in defining regional policy options while capturing economies of scale and scope from a regional approach. Regional approaches help where countries are too small or problems too complex to maintain adequate national capacity to address the weak evidence-to-decision link on their own. Some regional collaboration has worked well on the generation of evidence for agricultural policymaking as in the cases of ReSAKSS and ReNAPRI. Government-to-government dialog within regional organizations suggests the possibility of capturing some economies of scale and scope for national agricultural policy formulation based on evidence generated through regional networks. However, staff in national ministries still require the capacity to participate actively and to advise national policymakers on the results of regional collaboration. Success should be measured by contribution and influence rather than policy outcomes. The motivation for providing evidence should be understood to inform the key agents in the policy process, rather than to guarantee a particular outcome. The metric of success should be the contribution of the 18 research to informed debate, rather than attribution of an outcome to research products. Contribution can be tracked through metrics of awareness of research findings on the part of participants in specific policy deliberations, as well as through exposure in media. Contribution analysis is relatively new in the field of evaluation, but is recognized as appropriate for assessing the impact of policy analysis. Finally, greater support for the use of trusted agricultural policy analysis in decision-making is likely to yield very large net benefits. However, as in the well-established case of persistent underinvestment in agricultural policy research itself, underinvestment in the mechanisms for delivering evidence- based analysis at the right time and place in a format to inform agricultural policy making occurs because the large benefits are not fully perceived by those making the relevant budget decisions.3 Reaping tangible returns from the substantial efforts to date to improve data and policy analysis in African countries requires investment in the use of the product for the purposes intended. To address this, any effort to promote agricultural data collection, policy research, or policy analysis should be clearly linked to a process that ensures attention to issues of the highest priority to key policy actors. Further, specific provision should be made for linking analytical output to the timely provision of briefing materials required by decision-makers. In conclusion Agricultural decision-makers in Africa face a tightening environment for accountability to their constituents. Competition for resources is high and rising, as are the complexities of the agricultural policy agenda and the demands on decision-makers. Informed policy debates will require improved institutional capacity of public agencies and civil society organizations to curate, interpret, consolidate, refine, and use evidence. African countries are not uniformly positioned to move on this agenda, but some are clearly ready to do so, as suggested in the summary of cases. The participants in a companion consultation workshop whose results are summarized in Annex I stressed the importance of political will, or the degree to which a government actively seeks to embed evidence in the policy process. Government entities demonstrating strong political will should be favored in the allocation of logistical and financial support to move ahead. Shared funding would also be consistent with strong local ownership of the evidence cycle. 19 1 20 Challenge: The Weak Link from Increasing Available Evidence to Increased Use of It Agricultural policymakers in Africa and elsewhere face new challenges. Adaptation to climate change, adjustments in international and regional trading regimes, the acceleration of poverty alleviation, the integration of lagging regions, the creation of better jobs for youth in agriculturally-based economies, and the servicing of rapidly changing urban and regional food markets are just a few. These new demands increasingly burden the policy processes that need to address these challenges beyond the traditional concerns of government activity in agriculture. Agricultural policymakers increasingly need to prioritize across expanding decision domains and coordinate across an expanding set of actors and institutions, typically without a commensurate increase in resources. They need to anticipate the likely range of consequences of actions on a broader set of objectives. Heightened demands raise the need for careful consideration by government of policy options based on timely, relevant, and evidence-based analysis delivered to decision-makers in forms they can use and that they trust. Data to support informed agricultural policymaking in Africa have expanded rapidly in recent years, as have research knowledge, the tools to analyze data, and the number of trained people equipped to do so.4 The increasing need for informed policy choices and the capacity to generate evidence would seem to create an attractive natural match. We should be witnessing a steady flow of interaction between African policy analysts and policymakers, leading to well-informed policies. Instead, instances in which policy outcomes can be traced back to consideration of the evidence are the exception rather than the rule. Insufficient consideration of hard evidence in agricultural policymaking prevents adequate responses to complex new challenges. It is also a barrier to the continent’s expressed desire for greater self-determination in the pursuit of shared prosperity. 21 In contrast to greater reliance on external assistance in the past, most countries in Africa now largely use their own fiscal and forex resources to fund needed agricultural public expenditures. To match the increased reliance on domestic funding, policymakers both need and want greater access to locally generated advice and analysis that they can trust. Furthermore, remaining external assistance is derived from a large and diverse group of funding partners with their own priorities and constituencies. The diversity of partners often leads to mixed signals in the dialogue accompanying the assistance and reinforces the need for local capacity to manage these debates. A rising domestic investor class and active foreign direct investment (FDI) from a broadening pool of investors further complicate the funding and policy landscapes. Strengthening the whole agricultural policy value chain involves many discrete activities: building human capital, assembling better data, building research systems to use the data, fostering skills to interpret research findings, adding capacity for evaluation and impact assessment, enhancing communication and outreach, connecting with decision-makers in a timely manner, and monitoring of outcomes. Many elements of the policy chain or ecosystem are in place or are currently being strengthened. The present inquiry focuses on a single weak link: the better integration of growing African analytical capacity for agricultural policy analysis and its data-based outputs (“evidence”) with timely delivery of evidence-based alternative options to policymakers in a time frame and in forms that they can use. A number of models have been employed for decades to address this weak link. The capacity to use hard evidence in transparent agricultural policymaking received significantly increased attention after the food crises of the early 1970s and especially during what has been called the “Structural Adjustment” era, running from the early 1980s to the mid 1990s (Commander, 1989; Sahn et al. 1997). Cumulative fiscal and trade imbalances led to reluctance of MoEFs and external public donor agencies active in Africa to continue to engage in business as usual under these conditions. This drove substantial reforms in agricultural policy that have continued into the new century. The design of reform agendas required agricultural sector analysis and evidentiary input into national decision processes. The various experiences then and since in providing analysis and evidence to participants reflect a range of contexts, approaches, objectives, and outcomes. 22 One explanation for the lack of evidence used in agricultural policy making in Africa is that the decision-makers concerned are motivated by something other than promoting growth with prosperity, alleviating poverty, or increasing resilience to shocks. Examples here include seeking rents from decisions (the worst case) or, more frequently, building political constituency through purely distributive policies (Bates, 1981). Examples of such cases can be found in many countries, including some in Africa. The present paper however starts from the premise that the majority of agricultural policymakers in African countries today are in fact very concerned with efficiency and effectiveness in promoting growth, alleviating poverty, and increasing resilience. As those governments seek to strengthen their policy processes today while facing challenges both old and new, lessons from earlier experience on how to increase the evidence-base of agricultural policymaking will be informative. This paper provides a brief discussion of various models over the past 40 years for promoting the use of evidence-based policy analysis by decision-makers. The models chosen for discussion are selective and favor different experiences for which written assessments are available. The underlying economic and political contexts for agricultural policymaking vary across time periods and countries. Reasons for believing that agricultural policymakers in the region are under increasing pressure to base their decisions on analytical evidence are addressed in the next section. 23 2 24 Why Is Relevant Evidence More Important for Policymakers Now Than in the Past? What Has Changed? 2.1 u The rising share of agricultural public expenditure from domestic sources Although donors remain committed to supporting agricultural development in Africa, their relative importance in funding of agricultural public expenditures in Africa has decreased in recent years, as is illustrated in Table 1. The donor share remains large—but still reduced—in countries with significant civil unrest in rural areas in the period covered. It is quite modest in less fragile countries. Donor shares in Burundi, Mali, and Mozambique are on the high side, but elsewhere average around 15%. Broader surveys suggest that this downward trend in the share of donor financing still applies when off-budget foreign assistance to agriculture - still favored by some major donors - is included (Mink, 2016; Lowder and Carisma, 2012). Even as the overall share funded by MoEFs increases, at present most African governments are still under-spending on agriculture relative to other parts of the world on a share basis relative to total public expenditures and even more so on an absolute basis. This is the case despite the documented need for rural public services, infrastructure, and agricultural technology to sustain rural and agricultural transformation and adaptation to climate change (Goyal and Nash, 2016). 2.2 u The rising need for accountability in an era of tightening domestic budgets As financing from domestic tax receipts increases, so do calls for accountability that funds are being well spent. Although it remains well below the crisis levels that triggered the Heavily Indebted Poor Countries (HIPC) debt reductions, 25 government debt as a share of gross domestic product (GDP) is rising in many African countries (Onyekwena and Ekeruchwe, 2019; Gill and Kenan, 2018). Figure 1 shows that many countries are at or are close to the debt threshold recommended by the IMF. Moreover, the composition of debt has shifted. Table 1: Share of Agriculture Public Expenditures Financed by Donors 2007 2015 Burkina Faso 70% 14% Burundi 93% 76% Kenya 12% 0% Malawi 18% 6% Mali 71% 41% Mozambique 57%a 27% Senegal 54%b 36% Tanzania 36% 18% Uganda 53% 9% Source: African countries in the FAO/MAFAP database excluding Ethiopia and Ghana where comparability of the data requires further investigation. The narrow definition of public expenditures specific to agriculture is used (i.e. not rural roads, etc.). Accessed May 23, 2019 at http://www.fao.org/in-action/mafap/data/en/ Notes: (a) 2009; (b) 2010. Pressure for improvement of the composition of agricultural spending in Africa will increase as domestic resource pressures mount and accountability becomes more stringent (Mink 2016). Many countries overspend on short-term subsidies to individuals and under-spend on longer term agricultural research for all. Thus, public funds displace spending that could be done by the private sector, while needed public goods and services languish. Furthermore, the balance between capital and recurrent expenditures is often skewed, with diminished returns to each. Finally, budget execution is weak in many cases. Agricultural budgets are very often under-spent at the end of the fiscal year, jeopardizing new allocations in times of tight resources (AGRA 2018, Mink 2016). As a result, MoAs are increasingly under pressure to show to MoEFs and executive authorities both the rationale behind proposed spending and the impact of past spending, particularly in terms of net estimated impacts on 26 Figure 1: Government debt as a share (%) of GDP for African countries, 2017 Angola Benin Botswana Burkina Faso Burundi Cabo Verde Cameroon Central African Rep Chad Comoros Dem. Rep. Congo Congo Rep. Côte d'Ivoire Equatorial Guinea Eritrea Eswatini Ethiopia Gabon The Gambia Ghana Guinea Guinea-Bissau IMF threshold Kenya AMCP threshold Lesotho Liberia Madagascar Malawi Mali Mauritius Mozambique Namibia Niger Nigeria Rwanda São Tomé & Príncipe Senegal Seychelles Sierra Leone South Africa South Sudan Tanzania Togo Uganda Zambia Zimbabwe Sub-Saharan Africa 0 30 60 90 120 150 Source: IMF, 2018. Regional Economic Outlook reproduced in Onyekwena, C. and Mma Amara Ekeruchwe, “Is a debt crisis looming in Africa?” Brookings Institute, Africa in Focus, April 10, 2019. 27 forex and fiscal balances. The latter are not directly observable as they require some form of modeling to capture second and higher order adjustments. Such pressures may well increase further depending upon the global outlook for commodity prices and recession. Agreed regional and global accountability frameworks such as CAADP and the SDGs also add new reporting requirements for national policymakers. These focus additional public and political scrutiny on policy and funding prioritization. MoAs must also increasingly coordinate and broker priorities with a broader set of government institutions and private sector actors, such as ministries of trade and industry and trade associations. 2.3 u Given increased need, why is evidence underused in agricultural policymaking? There are several possible responses to the question of this sub-section. Some reasonable supply-side hypotheses about the lack of institutional capacity in MoAs to consolidate and translate existing evidence-based policy analysis for decision-making follow. First, agricultural policy systems in Africa are heterogeneous and complex, arguably more so than in the case of most other sectors of the economy. Complexity derives from the breadth of incidence of agricultural policy. That is, the high share of population—more than 2/3 on average—and geographic dispersion of people dependent on typically low agricultural incomes. Further, there is a diversity of institutional forms (such as parastatals) and tasks involved (Goyal and Nash, 2017). A study of public spending on agriculture in Mali from 2010 to 2015 noted 43 agencies supported by an agriculture budget accounting for 15 percent of all public spending (United Nations FAO/MAFAP, 2017). In part, difficulties in using available evidence-based analysis in agriculture decision-making reflect this structural complexity. Second, MoAs require a new mix of hard and soft skills to successfully integrate evidence into policymaking. Although more young Africans than in the past have the training needed for policy analysis and interpretation, few of them work in MoAs. MoAs have difficulty attracting and retaining staff with the relevant skills in analysis, communications, and project management. Third, ministries lack the incentives and systems to accelerate performance. Often, civil service rules require ministerial staff to move to new assignments 28 on a regular basis—limiting the impact of training or skills building. Incentives are limited (e.g. salary and recruiting rules, tenure-based promotion systems), and few countries use performance management systems at either the staff or project levels. Fourth, organizational culture and the political economy may not reward the use of evidence in policymaking. MoAs have been associated traditionally with politically-based campaigns and grand rhetorical pledges, rather than with technical analysis. This approach to policymaking has been costly in the past and is likely to be even less tenable in the future as the technical demands from MoEFs increase. Fifth, agricultural policy requires multi-stakeholder coordination on a larger scale than most mega-sectors. Decision-makers in other ministries, donor agencies, NGOs, and the private sector may be reluctant to follow the lead of MoAs in what they consider their own core areas. Beyond the five gaps identified above, a large range of other factors can be shown to feed into explaining why certain agricultural policies are decided on the basis of evidenced-based analysis and others are not.5 Cumulatively, these barriers may ultimately limit ministries’ abilities to effectively negotiate for funding, and develop, justify, and coordinate the implementation of policies to drive inclusive agricultural transformation. Unsurprisingly, a wide variety of institutional models have been employed over time in Africa to increase capacity to supply agriculture decision-makers with timely and evidence-based policy options and analysis. The nature of these models depends very much on the changing context of the time period when they first emerged in the independent countries of Africa from 1960 through the 1990s. The next section examines key elements of changing context that shaped the models that emerged. This is followed by an examination of a selection of agricultural policy support models that have been prominent in Africa in recent years. 29 3 30 The Evolving Context: 1960 to 1990 3.1 u Addressing the educational preconditions Agricultural policy research, outreach, and empirically grounded policy debate rely on a series of core building blocks: trained agricultural economists, good data and methods, communication skills, dissemination outlets, and institutional conduits for engagement with policy stakeholders. During the early independence years, a series of large-scale investments in African agricultural universities provided vital, long-term investment in human capital and an institutional architecture favorable for fostering productive and constructive agricultural policy debate (Eicher, 2003, 2006, 2009). Beyond creating the conditions for evidence-based policy research, university-to-university links evolved in some cases into vehicles that helped curate, consolidate, and communicate agricultural policy options and implications in a timely manner to the policymakers concerned. Stepping back in time, the context of the 1950s to the early 1970s included the Cold War, successful independence movements in developing countries, and East-West competition for the minds and allegiance of the emerging elites of new countries. Africa was no exception: in most cases independence was achieved as late as the early 1960s compared to much earlier in Latin America and Asia. The United States was the dominant public sector development financier in the heyday of foreign assistance following 1960. In agriculture, the U.S. model of agricultural development of the 1960s and 1970s was heavily influenced by the success of the Green Revolution in Asia (Mellor, 1998; Eicher, 2003). This emphasized a supply-side approach to boosting the productivity of cereals, including agricultural research, extension, and input supply systems (Ibid.). The availability of large-scale funding from USAID, and to a lesser extent channeled through the World Bank, permitted the launch of a series of African agricultural universities and research institutes across the continent. These were inspired by the U.S. land grant system and similar—sometimes earlier—replication of these models in Asia. 31 Early African examples included: Alemaya College of Agriculture, Ethiopia (now Haramaya University), founded with faculty support from Oklahoma State University from 1954 to 1972; Amadou Bello University at Zaria, in Northern Nigeria, founded in the early 1960s with long-term technical support from Kansas State University; and Malawi’s Bunda College of Agriculture, founded in the 1960s (Eicher, 2006; Goldsmith, 1990). Though experiences varied across settings, these early institution-building efforts shared a common commitment to collaborative, on-the-ground research and analysis involving international and local faculty, student training at the bachelors, MSc and PhD levels, and active support for extension programs in agricultural and social sciences. In these early years, agricultural universities and agricultural research institutes in Africa were almost always run by agricultural scientists, often with a plant breeding background. This was also the administrative structure for the international agricultural research centers supported in developing countries, including in Africa, by the Consultative Group on International Agricultural Research (CGIAR) starting in 1972 (Özgediz, 2012). The role of economists in these institutions tended to be limited in the early days to studies of the economics of technology adoption, or sometimes just “crop budgets” (Ibid.). A key part of the Asian model applied to Africa required strengthening public sector agricultural research institutions to add new sections that would focus on policies to complement the introduction of high-yielding varieties (Mellor, Delgado and Blackie, 1987). Agricultural policy research in Africa in the 1960s and 1970s tended to be limited to politically non-controversial issues of how to promote technology adoption or increase fertilizer use (Staatz and Eicher, 1998). Optimism about an independent future and rapid growth prevailed in Africa in the 1960s. In agriculture, cropped area doubled with the rapid expansion of cash crops after independence in many countries (Delgado, 1998). Peanuts, cotton, cocoa, and other commercially demanded crops for processing had higher returns than the predominant local food crops (Kherallah et al. 2002). They also had quite different labor profiles from food crops across the agricultural season. Since aggregate farm production in Africa in the 1960s was primarily constrained by labor availability for food crops during peak periods (such as the second weeding of cereals in the Sahel), new demands could be met by new supply without cutting back on other products (Kherallah et al. 2002). By 1970, in contrast to what would happen soon thereafter, aggregate food 32 supplies outside conflict zones were secure in Africa’s agriculturally dominated countries, real exchange rates were in balance, and the macroeconomic situation was generally strong (World Bank, 1981; Delgado, 2011). 3.2 u The transformative 1970s: the oil shock of 1973, food crisis of 1974, and expansion of foreign assistance to African agriculture The late 1960s—which witnessed several massive famines in Asia—saw considerable discussion in popular literature of coming famines in the developing world.6 Also, the early 1970s were a period of global instability in grain markets. Although the causes were multiple, massive grain purchases (more than 10 million tons from the US alone) by the U.S.S.R. in 1972 following crop failure were associated at the time with massive global grain price increases (Peters, 2009). Global grain and oil prices spiked (with the latter quadrupling) in 1973/74, draining forex reserves across Africa and making it more difficult to import grain (Delgado, 2011). Coincidentally, the Sahelian band of Africa stretching from Senegal to Ethiopia experienced a devastating series of crop failures from 1972/73 through 1974/75, leading to an estimated 100,000 deaths from starvation and disease exacerbated by undernutrition (von Braun et al. 1999). Inevitably these events affected attitudes in many African countries towards agricultural policies and decreased willingness to rely on trade for food security (Kherallah et al. 2002). The combination of local crop failures and high international prices also led to a major ramping up of foreign assistance to African agriculture after 1975 that continued until a peak in 1989 (Eicher, 2003). Beyond food and macroeconomic events in the 1970s, another set of issues began to change donor and national attitudes towards agricultural policy analysis and its use in Africa. The success of the Green Revolution in the 1970s in Asia and Latin America fostered a shift within the development community away from highly focused attention on hunger and famine toward a more generalized concern about poverty and related issues, including nutrition and gender equality (Mellor, 1998). A consideration of African issues in the 1970s could also justify a focus on poverty, nutrition, and gender, although aggregate food supply remained a serious problem in many countries. Subsequent research would show that as of 2004, three quarters of the absolute poor in 33 Africa were still in rural areas and derived their livelihoods largely from smallholder agriculture (World Bank, 2007b). The global shift in development priorities in the 1970s became known as a Basic Human Needs (BHN) focus, a name coined from a series of International Labor Organization (ILO) missions to Africa and Asia. BHN concentrated on finding ways to meet the needs of the poor in health, nutrition, and education as quickly as possible (Delgado, 1998). A corresponding shift in emphasis to rural livelihoods and access to services came at the expense of a concentrated effort to boost agricultural productivity and sectoral growth more broadly (Staatz and Eicher, 1998; Mellor, 1998; Eicher, 2003). This effect was aggravated by two related differences that distinguish the main period of donor support to agriculture—including policy-related—in Africa since the 1974 food crisis from earlier efforts in Asia. First, a multiplicity of donors were anxious to make their own mark in the same countries (Eicher, 2003), and second, support became more obviously projectized to keep better track of the relationship of outputs to a specific set of inputs (Lele, 1975; Delgado, 1998). With regard to the first issue, funding proliferated rapidly and outstripped the ability of national counterparts to coordinate projects dispersed across rural areas (Mellor, Delgado, and Blackie 1987; Eicher, 2003). Donors during the period emphasized cross-sectoral, geographically-defined projects, with flags of different donors flying metaphorically over different parts of host countries (Lele, 1975). National governments had to provide counterpart staff to each donor, which stretched demands on the scarce professional expertise of ministries and hindered the growth of a critical mass of qualified specialists within central ministries (Ibid.). Weak donor coordination and the projectization of aid from donors working on different agendas arguably undermined outcomes for many years after the 1970s. Before a reform of the CGIAR system from 2008 to 2011, for example, there were 57 independent project offices in Tanzania maintained by CGIAR centers (CGIAR Independent Review Panel, 2008). Bilateral donors similarly projectized and fragmented their assistance. Hindsight illustrates that concurrently with these developments in sectoral assistance, profound changes affecting agricultural incentives proceeded outside agriculture and to a large extent outside Africa in the 1970s. Many stakeholders—including foreign experts—dealing with African agricultural 34 policy issues had very different understandings of these extra-sectoral developments as they unfolded.7 That agricultural policymakers were unprepared for the big shifts of the 1970s illustrated the need for African MoAs to avoid a replay of unpreparedness for the challenges of the decades ahead. The crux of the issue came from a sudden and fundamental redistribution of global wealth in 1973 from industrial countries to petroleum exporters. This led to a game change in global lending, referred to at the time as the petro-dollar boom (Lütkenhorst and Minte, 1979). Foreign public and private lending for urban construction and services in Africa rose steeply, seemingly under less stringent financial conditions than had prevailed previously (Ibid.). The share of Africa in all foreign direct investment to all developing countries was a fairly constant 27% from 1970 to 1972, rose to 65% in 1974, and fell to 4% in 1980 (United Nations Conference on Trade and Development, 1999). At the same time, African real exchange rates appreciated substantially in part due to earnings from more valuable mineral exports and in part due to expansion of government spending on services and other non-tradable items (Krueger et al. 1991). The spending led to a brief boom in Africa in the mid 1970s as investment inflowed to cities and as agricultural exports increased with global economic recovery from the 1973 oil shock. The stagnation of Africa’s agricultural exports later in the decade and the need to repay loans for urban investments turned the boom to bust and left a legacy of debt (Krueger et al. 1991; Kherallah et al. 2002; Delgado, 2011). 3.3 u “Structural Adjustment” in the 1980s and 1990s A devastating Africa-wide fiscal and balance-of-payments crisis came to a head in 1979. This in a region where agriculture at the time provided 85 percent of employment and more than half of exports, and foreign donors financed more than half of public expenditures in agriculture (World Bank, 1981; Delgado, 1998; Mink 2016). Agriculture could not be exempted from needed changes in policies. At the urgent request of African MoEFs, the World Bank produced a diagnosis of Africa’s economic crisis at the time, the 1980 Berg Report (World Bank, 1981). This was primarily a macroeconomic policy story of inappropriate sectoral incentives in African countries. According to the report, civil service expansion and urban investment had led to a need for high agricultural 35 taxation.8 Over-borrowing for urban construction and services during the go-go years following the petro-dollar boom after 1973 had also led to the need for fiscal and forex resources for debt service that only agriculture at that time could satisfy. Both phenomena led to overvalued real exchange rates and a system of overall incentives that taxed farmers—the main producers of exportable goods—explicitly and implicitly and favored urban sellers of services and consumers of imported goods. Political economy inquiry at the time also stressed the relative lack in Africa of politically influential rural notables with agricultural interests compared to Asia and Latin America. This resulted in urban-bias in the civil-service oriented political economy in Africa (Lipton, 1977; Bates, 1981). Regaining macroeconomic health in Africa would fundamentally require “getting (relative) prices right,” especially in tradable agriculture (World Bank, 1981; World Bank, 1986; Commander, 1989). In many countries, this was thought to require liberalizing markets for export crops (significantly raising prices received by farmers, but reducing government revenue) and shrinking the civil service (Ibid.; Kherallah et al. 2002). Thus, foreign donors began to concentrate on promoting incentive policies in agriculture that justly rewarded mainly agricultural producers of exports by removing policy distortions that taxed them while also removing policies that created artificial subsidies to mainly urban consumers of imports (Delgado, 1998). Both government and donor interest in agricultural policy issues in the few years following 1980 shifted from things such as technical discussions of fertilizer use and seeds to a strategic and politically controversial, sometimes ideological, debate on how to re-distribute incentives throughout the economy (Krueger et al. 1991; Mellor, 1998; Delgado, 1998; Jayne et al. 2002). In contrast to the World Bank’s diagnosis in the Berg Report, the Organization of African Unity’s Lagos Plan of Action for the Economic Development of Africa 1980- 2000, written at approximately the same time, largely attributed the region’s ills to excessive openness to global economic shocks and misconceived donor policies (Organization of African Unity, 1980). In contrast to current overall policy debates in Africa in which agriculture is rarely featured prominently, both reports focused on key, but very different roles for agriculture in the overall development process.9 Ideology held sway and governments had little recourse to assessments of competing positions through analysis backed up by empirical evidence. 36 Thus, from the economic policy standpoint, the Structural Adjustment era in Africa was an unusual time of alignment of the big public creditor countries through the informal Paris Club. The World Bank and IMF worked with MoEFs to resolve macroeconomic problems. Donors and MoEFs proposed adjustment actions in all sectors and in the overall macroeconomy, backed by the analytical capacities of the Bretton Woods institutions and creditor nation treasuries.10 MoEFs coordinated line ministries that oversaw sectoral actions. Many African MoEFs had difficulty at the time formulating and presenting sectoral policies without adequate backup from the line ministries concerned, especially in the case of agriculture.11 Donors and some MoEFs voiced frustration with a lack of sectoral responses to what was seen as the indisputable need to address balance of payment crises through re-alignment of macroeconomic incentives. Relatively little had changed in agricultural policy in many African countries by 1983 (World Bank, 1986), unlike the case in Latin America and Asia.12 This led to the start of application in 1984 of requirements for governments to satisfy an increasing number of formal policy conditions—especially in agriculture—in order to access budget support, a practice termed conditionality that was supported by the World Bank, the IMF, and most major donors (World Bank, 1986; Commander, 1989; Sahn et al. 1997). Agriculture was not the only sector at issue, but as in global trade policy at the time, agricultural policy was one of the most contentious. External donors made it clear that more-evidence based policy making in agriculture was required for their continued financial support (Commander, 1989). An example of the difficulty of the times is furnished by the Structural Adjustment Loan/Credit (SAL I) signed by the World Bank with Burkina Faso in 1991, the first one signed with that country (World Bank, 1991). It provided US$80 million from IDA funding, with a supporting amount of US$92 million from the EC, AfDB, Canada, France and Germany (Ibid.). It took nearly two years for the Burkina MoEF to negotiate with sectoral ministries and WB missions, after years of resistance to some of the conditions from sectoral ministries as acknowledged in the SAL document. The latter spells out over 100 specific policy-related actions, many of them sectoral, that the government would need to undertake as its side of the bargain to access the funds (Ibid.). Much of this painful process for both ministries and donors likely could have been mitigated in agriculture by a proactive, internal, national process of analysis and decision-making informed by evidence. 37 4 38 Capacity-Building for Agricultural Policy Analysis and Advice Since 1980 Donors funded efforts to promote the generation of evidence for agricultural policymaking and to support institutions that housed that capacity took various forms from the early to the mid-1980s onwards. These were influenced by national differences and needs and by contextual factors such as the state of the national agricultural education and research system in 1980. 4.1 u The ascent of country-specific agricultural policy analysis units after 1980 Where MoEFs felt a need for better support on analysis of agricultural policy issues internally during the Structural Adjustment era, some encouraged donors to establish APMEs directly in MoAs. This form of institutional arrangement responded to the view at the time that public entities could and should occupy the center of the policymaking and implementation monitoring chain. This helped enforce discipline in tasks taken on by the unit. It also provided institutional buy-in and memory within government for what was achieved. The next section discusses a well-evaluated case from Tanzania. In some countries, U.S. universities had already been funded to work with national universities on producing national agricultural policy analysts since at least the 1970s. In the 1980s and 1990s, national staff with PhDs were returning to lead efforts. It was natural that distinct agricultural policy analysis institutions would be established in those countries in affiliation with the national universities (Idachaba 1996; Eicher 2009). These units had varying degrees of autonomy vis-vis the university in question. Factors influencing the relationship included: requirements for multidisciplinary teams to approach certain topics, ease of unit access to external funds on a specific set of topics, staff receiving higher than Ministry of Education scales or consultancy payments, and a reduced the level of central university overhead charged. 39 The Agricultural Policy Research Unit (APRU) at Bunda College in Malawi and the Tegemeo Institute at Edgerton University in Kenya are examined as case studies in the next section. Both were offspring of university-to-university collaboration. Both provided institutional vehicles through which research faculty and graduate students could contribute the empirical evidence necessary for informed agricultural policy debates. Besides the freedom to pay higher salaries and to retain more qualified staff, a key element was the freedom to tap into funding from a variety of donors. Some of these included agencies that had very specific asks, but that were not willing to contribute beyond the variable marginal costs of the output they wanted. Thus, they were “free-riding” on the coat-tails of other donors who were building the core of the institution (Jayne et al. 2019). Witnessing the difficulties MoEFs were having responding to World Bank and IMF economic queries in African countries during the Structural Adjustment era eventually led donors, including the World Bank, to support national efforts to establish macroeconomic policy think-tanks serving the needs of the MoEFs. These were largely separate from line ministries, but typically served MoEFs and sometimes other ministries under a special relationship.13 They would provide evidence-based policy analysis to national decision-makers that would strengthen their hand in negotiations with external creditors and in fact served both sides well. Not only have these institutions survived and broadened over time under national MoEF sponsorship in a number of countries such as Kenya, Senegal, and Uganda, but the economic think tank model is being replicated across Africa with a broader mandate than Structural Adjustment issues, although not without concerns over funding.14 The general economic policy think-tanks serving MoEFs in the 1990s tended to have small agricultural programs at the time. However, these agricultural programs have shrunk over time and are no substitute for in-depth agricultural policy research, analysis, and advisory activities on a par with what is needed for agricultural policymaking. In 2003, in response to coordination issues in funding and the multiplication of its own project offices that had been growing for years in several African countries, IFPRI commenced a series of Country Strategy Support Programs (CSSP) funded primarily by multiple projects and donors across a variety of topic areas in each country. These were now run as a coordinated whole, with varying degrees a of closeness to host country governments across countries. 40 This was a case of coordination of donors by the agency being funded in collaboration with the host country government. Part of the projectization of external public investment in African agriculture discussed above was the practice of individual projects establishing project implementation units (PIUs). PIUs for agricultural projects were embedded within MoAs in Africa. They typically included a few higher-level staff capable of project evaluation, reporting, and sometimes policy evaluation. For a small country with few agricultural projects, at least one of which was large and longer-term, the PIU for that project would in effect be a mini APME. This APME would oversee implementation of that project for the donor and the government as well as analyze agricultural policies and investment issues more broadly for the host country. Where the number of projects is large, such coordination needed to be provided by a different kind of unit that also answered directly to government in the same sense as a PIU. The solution adopted in Ethiopia was an independent government agency, the Agricultural Transformation Agency (ATA), that has the capacity to hire based on merit, pay higher salaries, and better retain key staff. ATA is discussed further as a case. It is an example of a national mega-PIU that coordinates across agricultural projects and program activities and ensures proper implementation. ATA is in effect one solution to the still largely unresolved issue of national sovereignty raised by the Structural Adjustment era. However, it is a solution better adapted to a large country with a large agricultural investment program, a high degree of external support to agricultural public investment (and thus interest in supporting ATA), a large supply of high caliber national skills in economics and agricultural policy, and tolerance of a high degree of government intervention in agriculture. As will be discussed, the replicability of this model in other countries in Africa is still to be determined. Finally, a relatively new country-level model for incorporating evidence into agricultural policy has surfaced recently in Senegal, where past capacity building efforts of development partners and domestic entrepreneurs have led to the creation of nine public and nongovernmental institutions carrying out various forms of agriculturally-relevant policy research and analysis. Little of this work was being used systematically for policymaking by the Senegalese MoA prior to the launch of PAPA project in 2016. PAPA is housed in the MoA with the mandate to communicate policy priorities and specific queries to the 41 network of nine domestic policy analysis institutions, facilitate their access to funding, and curate their output for policymakers. PAPA, which is discussed below, could be termed a networking model of policy support leveraging and aggregating domestic capacity. PAPA has been supported until recently by IFPRI and Michigan State University under USAID funding that appears to be ending. The question remains whether the benefits of PAPA can be maintained without external funding and analytical advice. 4.2 u Seeking external advice on agricultural policy without building new institutions Many countries may want technical input in a specific area such as an agricultural public expenditure review, but not want to support the capacity for such work nationally. FAO, among its other functions, provides on-call technical assistance in agricultural policy to individual countries that lack internal capacity in areas of high priority for agriculture. Such technical expertise is typically short-term and does not involve FAO staff taking up residence in the host country. Core funding for such missions has become increasingly problematic over time. The rise of substantial trust or project funding, compared to core funding, gives a more modern twist to the international agency model for providing specific on-demand services to developing countries. The Monitoring and Analyzing Food and Agricultural Policies (MAFAP) program has been trust funded by a number of donors since 2009 and housed within FAO. It is analyzed in a case study in the next section. MAFAP provides in-depth assessments of the agricultural incentive environments in selected countries, with an emphasis on Africa. Drawing on this background analysis, MAFAP offers assistance to MoAs that request it in the area of analysis of implementable options for food and agricultural policies. It is especially well known for agricultural public expenditure reviews and quantification of distortions in incentives. Again, MAFAP missions are short-term, with staff posted to FAO headquarters and FAO regional offices. In another model popular in the last decade, embedded foreign technical advisors are placed within MoAs. The embedded advisors address the need for increased country familiarity and proximity of interactions on policy advice. In the 1960s and even the 1970s in the smaller and poorer countries, it was more the norm than the exception that foreigners, typically paid by the United Nations or 42 a bilateral donor, ran the technical services of MoAs.15 By comparison, at India’s independence in 1947, an Indian scientist filled virtually every BSc level post in agricultural research. Only 10 percent of agricultural researchers in Africa in 1960 were nationals of their country of residency (Beintema et al. 1998; Eicher 2003). This changed over time as African nationals completed higher education abroad and returned to serve in their national governments. The embedded advisor model has recently resumed, specifically in the agricultural policy or agricultural project implementation coordination fields. Embedded advisors covering food and agriculture today typically function at a high advisory level, primarily on policy and strategic issues. Bilateral donors such as DfID, AFD, and USAID have such personnel in selected countries, and the EC has a major trust funded program in this area called FIRST. This allows the placement of senior technical-level FAO advisors in 12 African countries. The FIRST project is discussed in the next section. TBI, also discussed below, uses a similar model, with some differences. TBI advisors focus more on management and implementation and typically are embedded at the highest executive level. 4.3 u Regional approaches to improving the use of evidence in agricultural policy The African Union/NEPAD Comprehensive Africa Agriculture Development Program (CAADP) was launched in Maputo in 2003. It represented a new vision for agriculture’s role in Africa, a reassertion of African ownership of agricultural development and policy, and a high-level political commitment to take responsibility to provide resources to agriculture in ways that were explicitly to accelerate agricultural growth. Africa’s heads of state committed to CAADP within the broader political context of transforming the Organization of African Unity (OAU) into a new African Union (AU) as a more accountable and effective vehicle for managing Africa’s challenges and opportunities. It was intended to include a within-Africa peer review function that the OAU had never taken on. As CAADP has evolved since 2003, the role of evidence in policy and investment decisions related to CAADP has become increasingly important. This first involved the tracking of the CAADP financial commitment that African governments put at least 10 percent of their annual fiscal outlay into agriculture, and then--with significant help from outside partners--a much broader provision of evidence on agricultural financing, growth, and impact. 43 Over time and especially after 2005, CAADP became the main regional effort espoused by the African Union Commission and African countries for self- determination in the priorities in agricultural investment. CAADP contributes both to the will for supporting the generation of evidence and increasing its actual use in a manner owned by African MoAs. With the 2014 Malabo African Union summit, African governments manifested a renewed commitment to evidence-based policy and investment through CAADP. At that summit, Africa’s heads of state set out a broader and more ambitious set of food security targets in the areas of agriculture, nutrition, trade, environmental sustainability, and poverty to be achieved by 2025. They also directed the African Union Commission to report back to the AU summit every two years on progress achieved towards these 2025 goals, country by country. The result, with significant development policy research partner support, has been a set of transparent, publicly available indicators across almost every country in Africa feeding into a Biennial Review of progress on Africa’s food security (BR). The BR is designed to inform diagnosis of the reasons for and remedies to countries falling short of achieving the progress required to achieve the goals for 2025. While still being perfected, these indicators and the CAADP BR process have introduced a major new step in government-owned, evidence- based policymaking and investment decisions in the region. In policy analysis and advice, as in agricultural research more generally, most African countries suffer from a small country problem. This provides an additional impetus for regional approaches even beyond the political will for regional self-determination expressed in CAADP. Where needs exceed national capacity and are replicated across borders, regional networks and alliances can leverage competence and enhance results. Regional networks—beyond partnerships with non-African entities—may also increase national and regional ownership for results. Just as regional efforts in technical agricultural science have increased in recent years, regional policy analysis and research networks provide data and policy analysis to support national CAADP efforts. The Regional Strategic Analysis and knowledge Support System (ReSAKSS) and the Regional Network of Agricultural Policy Research Institutes (ReNAPRI), both discussed in the next section, are two different but 44 complementary approaches in this regard. ReSAKSS is a consolidator and translator of knowledge at the continental, regional, and sometimes national levels supported by IFPRI. It directly addresses knowledge needs for CAADP implementation with the explicit endorsement of the AUC and AUC-AUDA and the Regional Economic Organizations concerned. It responds to a rising interest in self-determination in agricultural matters in Africa (Brüntrup, 2011). ReNAPRI is a self-coordinated group of national agricultural policy research institutes in Southern and Eastern Africa that undertakes policy relevant agricultural research and analysis.16 The network was originally facilitated by the fact that Michigan State University had had prior individual partnerships with most of the constituent members, who now wanted to relate directly to each other. Activities associated with ReNAPRI do not directly fill the targeted gap in MoAs for capacity to consolidate and use available evidence-based policy analysis. However, they do facilitate access to proven African skilled researchers in this area for governments that wish to make use of them. 45 5 46 Case Studies Following on the narrative above, a number of cases provide more specific insights on what went well and what went less well from the standpoint of enhancing the capacity of ministries to consolidate, broker, and communicate a set of feasible policy options. The latter should include succinct assessment of advantages and costs detailed for different options for policymakers to choose amongst at the time they need to choose. 5.1 u Agricultural policy and monitoring units within MoA in the 1980s and 1990s: the Tanzania Agricultural Sector Management Project The operational model: The Government of Tanzania secured funding under a World Bank project over the period 1993-2001 to establish an agricultural policy unit within the MoA.17 An inter-ministerial steering committee, including the Ministry of Finance, oversaw project implementation. Funding of US$24.5 million was provided under the IDA credit. The policy unit provided technical assistance on specific tasks requested by the MoA, the Ministry of Finance, and the entity charged with implementing the privatization of state enterprises. The unit also supported training of MoA staff to undertake new functions as the economy transitioned from one of central planning to one with greater reliance on markets and upgraded agricultural statistics. The approach to targeting of evidence in policymaking: In the Tanzanian context at the time, targeting of evidence had to be preceded by institutional reforms to create space for the assessment of policy options. The project was designed to facilitate the transition of Tanzania’s agricultural sector from management reliant on central planning and administration to greater emphasis on markets. How policies operate to effect outcomes in a market environment required new understanding on the part of decision-makers. The approach thus included background analysis to support restructuring of the MoA, technical analysis of individual parastatal firms to guide accelerated privatization, and studies of options for reform in input supply, state procurement, trade, and the provision of services. Part of the technical assistance was commissioned directly, largely by the privatization entity, and applied directly in the ongoing privatization. Part 47 was of less immediate application and served as background for consideration of wider reforms in the sector. What worked well: The Tanzania Agricultural Sector Management Project (ASMP) benefited from the standard evaluative tools of the World Bank, including implementation completion reports by the World Bank and by the Government of Tanzania soon after closing (World Bank, 2001). It also was reviewed in a Project Performance Assessment Report after an interval of 6 years (World Bank 2007a). Key elements recognized as success included reorganization of the MoA; retrenchment and decentralization of staff; analysis to underpin the privatization of 80 agricultural parastatals, subsidiary companies ,and operating units; training of many MoA staff; and establishment of the agricultural statistics unit. The studies supporting practical decisions on privatization were particularly effective. The associated recommendations were regularly applied by the Government of Tanzania with planned effect. A critical mass of MoA staff participated in short courses delivered locally, short courses abroad, or degree training abroad. The intensity of training was such that prevailing attitudes within the MoA changed, and support for the reforms strengthened despite a backlash that could have been expected due to the retrenchment and decentralization of staff. What worked less well: The evaluative reports note several areas of weakness. The decision to work within and focus solely on capacity development of the MoA, instead of a broader or more independent effort, secured strong support from staff within the MoA. However, it also had a less desirable effect of subjecting new trainees to the vagaries of ministerial reorganizations and budget cuts. Continuity of provision of evidence over the longer run was undermined. The studies undertaken were judged to be of generally high quality, but not all led to applications in practice. Those that had the most direct practical impact were the feasibility studies for divestiture and privatization of the parastatals. Studies of equal importance, but with less urgent practical application, had less success in application. Examples were input supply, marketing, and food security strategy. These had less clearly committed clients, or clients facing less urgency in policy reform than the fiscal savings associated with the privatization that the Ministry of Finance targeted. The initially promising improvements in agricultural statistics fell victim to underfunding, with the result that later studies suffered from a paucity of data. 48 The MoA split into three parts and the Ministry of Finance retained the savings from privatization. The Government of Tanzania was legally bound to provide counterpart funding, but failed to sustain the agreed flows. Savings from privatization were not shared to fund improved staff incentives within MoA. The working environment for policy analysts within the MoA deteriorated. World Bank funding under the project ceased in 2001. Lessons learned: The experience with the ASMP project and the subsequent trajectory of capacity development for policy analysis in Tanzania highlights the importance of demand for evidence on the part of an empowered client. In this case, the Ministry of Finance needed help with privatization and sought technical guidance, which was applied. Similarly, the MoA faced an imperative to restructure and welcomed advice on how to do it. Although this was judged to be a successful project by the World Bank’s operational assessors and by the Independent Evaluation Group, it did not leave a legacy of long-term influence of evidence on Tanzanian agricultural policy (see, for example, Isinika et al. 2016). The creation of a policy unit did not assure institutional longevity in the form of protection from vagaries of ministerial budgets. The provision of evidence appears to have been too narrowly targeted to build the needed constituency within government as a whole and within civil society. In settings in which the policy process is more fully developed than in Tanzania in the late 1990s, both narrow targeting of messages (that is, to a specific and receptive client) and broader targeting are desirable to achieve practical impact and to underpin longer term support for the continued provision of evidence. 5.2 u Agricultural policy think-tanks or agencies derived from external university-to-domestic university programs 18 in Kenya, Zambia, and Mali Africa’s agricultural universities provided a foundation for agricultural policy debate by training generations of agricultural scientists, including the agricultural and natural resource economists capable of conducting empirically grounded policy research. They have also built extension and communication specialists and programs of applied agricultural research using scientific methods. Agricultural graduates staff Africa’s MoAs and over time have shaped the MoAs’ capacities to consolidate, broker, and communicate a set of feasible options for policymakers. To support these efforts, a long series 49 of donor-funded policy projects have enabled agricultural universities to support empirically grounded agricultural policy debates in many African settings. The examples below describe three Michigan State University (MSU) efforts in Kenya, Zambia, and Mali. Together, these examples demonstrate the multiplicity of institutional pathways through which investments in policy research leave behind enduring institutional legacies. The operational model for Tegemeo: Kenya’s Tegemeo Institute of Agricultural Policy and Development (“Tegemeo”) provides an example of how early investments in African agricultural universities have contributed to evidence- based policy debates over many decades. Tegemeo is a policy research institute under the Division of Research and Extension of Egerton University in Nakuru, Kenya. For over 20 years, from 1997 to 2017, faculty from MSU and Tegemeo designed and conducted a program of collaborative agricultural policy research. Although MSU served as the lead institution contracting with donors in the early decades, Tegemeo has subsequently become the prime contractor in negotiating with government and donors and leading major research initiatives. The Tegemeo approach to targeting of evidence in policymaking: Policy research under the Tegemeo Agricultural Policy Research and Analysis program aimed to inform and influence policies and development interventions affecting agricultural productivity, incomes, and food security. Together, Tegemeo and MSU researchers have produced hundreds of policy research papers built on rigorous empirical data collection. As Tegemeo took over the leadership of the partnership, it also became the principal conduit for providing analysis to decision-makers. What worked well with Tegemeo: The pairing of international researchers with local researchers to conduct joint policy research and outreach was effective in building local analytical and policy outreach capacity. The volume and pertinence of this stream of empirical research proved vital in establishing the credibility of Tegemeo as well as the utility of empirically rigorous policy research. The administrative and intellectual leadership shifted to Tegemeo once capacity was demonstrated and proved influential in policymaking due to increased ownership by African decision-makers. What worked less well with Tegemeo: During times of plentiful soft money, staff incentives and retention remain strong. Conversely, the ebbs and flows 50 of external funding led to periods of stress on local staff and compromised the institute’s ability to retain experienced staff during funding lulls. Lessons learned with Tegemeo: Tegemeo’s long experience suggests that national governments or other domestic stakeholders will need to commit resources to ensure stable core budgets that will enable the institutes’ leaderships to retain a critical mass of experienced researchers over time. The Operational model in Zambia, leading to IAPRI: Although most donor- funded policy projects leave little in the way of enduring institutional footprint, Zambia’s experience offers a welcome exception. The Food Security Research Project (FSRP) provided a 15-year funding vehicle, from 1999 to 2013, through which MSU and Zambian researchers conducted a stream of important agricultural policy research initiatives under the coordination of Zambia’s Agricultural Consultative Forum (ACF). The latter was a sectoral consultative platform managed jointly by the private Zambian National Farmers Union, the Zambian MoA, and agricultural donors. The stream of research produced by FSRP provided empirical evidence relevant to a range of highly visible policy issues affecting maize marketing, trade barriers, crop diversification, land consolidation, and input subsidy programs. Over time, a growing array of stakeholders came to appreciate the value of this empirical analytical work, motivating the establishment of a local research institute, the Indaba Agricultural Policy Research Institute (IAPRI) in 2011. Guided by a local board of directors, IAPRI operates as a non-profit Zambian company and receives basket funding from a variety of donors. Stakeholders believe that this basket funding helps to maintain the institute’s reputation for independence. The approach to targeting of evidence in policymaking in Zambia: Zambia’s FSRP and subsequently IAPRI have adopted several practices aimed at targeting critical information to key decision-makers. Through a formal advisory board–initially through the ACF, subsequently through the IAPRI board of directors–key agricultural sector stakeholders representing farmers, traders, and government help to identify research issues of particular concern. What worked well in Zambia: The project period served essentially as an audition, demonstrating to a wide range of stakeholders the value of independent, 51 evidence-based policy research. IAPRI’s reputation for rigorous empiricism and intellectual independence has, in turn, motivated an array of different donors to support the institute’s agricultural policy research agenda. This basket funding further enhances the institute’s credibility and independence. What worked less well in Zambia: Although born from university-to-university collaboration, IAPRI remains an independent policy think-tank due to concerns about the financial impact on both the institute and staff from closer integration with the University of Zambia. This raises the question of the long- term sustainability of its funding outside the Government of Zambia and of its continued outreach to policymaking circles (Jayne at al., 2019). Lessons learned in Zambia: Policy relevance matters, as does institutional leadership and independence. Regular stakeholder outreach coupled with a formal advisory board has enabled IAPRI to execute a highly relevant, forward looking research agenda. The shift from a single donor to a diversified array of financial backers for IAPRI’s policy research agenda has clearly enhanced the institute’s reputation for objectivity and independence. Rigorous empiricism has characterized FSRP and IAPRI research and policy analysis. Careful sampling and detailed primary data collection - through farm household surveys, market monitoring, and structured interviews with traders and consumers - have been hallmarks of IAPRI research. This reputation for scientific rigor has certainly helped to improve the credibility and acceptability of findings, even on contentious policy topics. The operational model in Mali: MSU faculty have collaborated with Malian colleagues since the mid-1980s on a series of research and capacity building efforts aimed at strengthening the understanding and institutional capacity for improving agricultural productivity and food security in the country. Initial work began with a consortium of government agencies and donors that proved instrumental in enabling the successful liberalization of Mali’s staple cereal markets in the mid- 1980s. Over time, a shifting array of foreign assistance partners–including USAID, EU, FAO, IFAD, AGRA and the Syngenta, Hewlett, and Bill and Melinda Gates Foundations–contributed financial support for these evolving efforts. The Malian model consisted of human capacity building, financial support, and technical assistance through MSU to address an evolving array of policy issues. 52 Over the course of multiple decades, these externally supported local collaborations led to the design of a series of new public agencies designed to respond to specific agricultural policy needs. The Observatoire du Marché Agricole (OMA), Mali’s price information service, monitors 77 agricultural markets throughout Mali. It was set up in a period of cereals market liberalization. The Commissariat à la Sécurité Alimentaire (CSA) implements Mali’s National Food Security Plan, including its early warning and response system, the Système d’Alerte Précoce (SAP). The Economie des Filières (Ecofil) is a new division of the Institut d’Economie Rurale (IER), Mali’s agricultural research institute. It studies agricultural marketing, agribusiness processing, farm input supply, and value chain dynamics in Mali. An agricultural economics and agribusiness program was added to the Institut Polytechnique Rural/Institut de Formation et de Recherche Appliquée (IPR/IFRA), Mali’s agricultural university. Today, the Malian government has integrated these policy units within government and has provided funding for the four local agencies, which continue to contribute to ongoing agricultural policy debates. The approach to targeting of evidence in policymaking in Mali: Targeting of evidence-based empirical research emerged in response to evolving public and donor policy concerns. In the early 1980s, the liberalization of Mali’s cereal markets served to focus research efforts. As government price controls dissipated, the need for careful monitoring of the emerging private markets became central to the Malian government’s political stability and their willingness to devolve price-setting authority to private traders. Intermittent droughts have similarly triggered concerns about early warning systems and the institutions that monitor and report to government decision-makers. What worked well in Mali: OMA’s early reports contributed to the confidence required to liberalize cereal markets in Mali. Subsequently, they have provided ongoing critical information help to shape food policy. Examples include supporting discussions at meetings of the Council of Ministers, leveling bargaining power among cereals market participants, and providing objective information to banks to value cereal inventories pledged as collateral for loans. Early efforts to train Malian scientists have borne fruit as a generation of agricultural economists trained at MSc and PhD levels in the 1980s and 1990s now assume leadership of key research, government, and private sector institutions. What worked less well in Mali: Government commitment to fund these new agricultural policy institutions has bumped up against serious overall public 53 resource constraints. Throughout the Malian civil service, modest salary incentives coupled with aging workforce have led to a steady drip of retirements of most experienced Malian scientists. Lessons learned in Mali: Continuously evolving policy pressures have dictated Mali’s analytical and institutional needs. Mali’s policy institutions have responded, and they will undoubtedly continue to adjust further over time. Early capacity building, primarily through international training of over three dozen MSc and PhD agricultural economists, has animated Mali’s policy institutions over the past generation. But this early wave of scientists is now retiring. New issues such as climate change and food safety under higher agricultural input intensity demand new skill sets and the renewal of human capital. 5.3 u An independent agricultural think tank linking universities, the private sector and government: The South African Bureau for Food and Agricultural Policy The operational model: The Bureau for Food and Agricultural Policy (BFAP) is an independent, not-for-profit company established under South African law, similar in other settings to a think-tank. BFAP was established in 2004 as a virtual network that consisted of researchers associated with the University of Pretoria, the University of Stellenbosch, the Provincial Department of Agriculture in the Western Cape, and the private sector (Traub, 2012). BFAP’s mandate is to inform decision-making by stakeholders in the agro-food, fiber, and beverage sectors of Africa and to develop capacity in analysis and research relevant to the sector. Initial funding consisted of a combination of direct financial contributions by institutions (the Maize Trust, ABSA bank, Wine Tech, amongst others) and in-kind contributions through the secondment of staff, access to databases, and models. BFAP works with clients and partners in the public, private, university, and NGO/donors sectors. BFAP stands out in respect of its strong links with the private sector and farm groups. A memorandum of understanding (MoU) exists with the Department of Agriculture, Forestry and Fisheries (DAFF) of South Africa and with the National Agricultural Marketing Council (NAMC). External partners include: the Food and Agricultural Policy Research Institute (FAPRI) of the University of Missouri, FAO, OECD, the Thünen Institut, and ReNAPRI. 54 The approach to targeting of evidence in policymaking: BFAP responds to specific requests by key agents in the policy process, often working under contract to deliver agreed outputs. These inform the positions of participants in policy debates. BFAP also releases regular products addressing forecasting, scenario planning, commodity market modeling and outlook, value chain analyses, food prices, and farm-level economics. These products receive due attention in the media and serve as background documents for discussion of agricultural policy issues. What worked well: The BFAP think tank model has several key features that contribute to success. The staff are recognized to be highly competent, with a range of disciplines represented and good engagement with relevant counterparts nationally. Several staff maintain international visibility and participate regularly in international fora. BFAP is thus not exclusively inward- oriented and is positioned well to interpret the changing global context for South African policymakers and business leaders. The combined mandate for policy and business analysis heightens the visibility of BFAP in the policy and agricultural business arena. Issuance of regular products such as outlook reports and active engagement on public-private sector platforms including the Maize Forum and the Fruit Industry Round Table provides predictability and facilitates access for policymakers and organizational leaders who need input quickly. Visible engagement with stakeholder groups and organizations such as the South African Agricultural Economics Association provides opportunities to interact with academics, public officials, producer organizations, and representative of private firms. MoUs with key public agencies exemplify trust and facilitate engagement in the policy process. The linkage with ReNAPRI and close relations with The University of Pretoria and The University of Stellenbosch and their teaching programs for students from throughout southern Africa give regional exposure. Because BFAP is independent and outside governmental structures, it is little affected by electoral cycles or vagaries of public budgets. As can be seen from Figure 2, many of the directly commissioned studies on specific policy issues have led to actions being either adopted or implemented. The impact of the background studies is not shown in the figure since their contributions are likely to be more diffuse and would require application of contribution analysis and surveys of policymakers. 55 Figure 2: BFAP policy engagements over 15 years and policy outcomes Stakeholder Policy Policies Time/Duration Outcome Request Recommendation Competitiveness, Year: 2003 Wheat crises Wheat Forum breeding, grading Period: 3 months classification Food Price Year: 2003 N/A N/A review Year: 2005–’07 ABSA bank, Blending rates, Biofuel policy Period: 6 months DTI,IDC,NAMC impact analysis Year: 2009 WTO third party - Maize WTO Maize Trust Period: 4 weeks affected Thailand Year: 2010 Impact on growth & SATI market closure Period: 2 months jobs, market access Scenario analysis: Wheat tariff N/A DTI & NAMC Impact on producers & consumers Year: 2010 National Planning NDP Chapter 6 Period: 8 weeks Commission Year: 2012 Government/Labor/ Minimum wage Minimum wage Period: 2 weeks Industry increase 50% Impact of citrus Year: 2013 Impact on growth CGA & DoA WC black spot Period: 6 weeks & jobs Chicken tariff Year: 2014 Alternative scenarios DTI & NAMC policy Period: 8 weeks for tariff levels Impact of mining Year: 2015 Impact on jobs, GDP, Maize Trust on agriculture Period: 3 months food prices Year: 2015 AGOA negotiations Chicken trade agbiz Period: 8 weeks & anti-dumping Importation of white Early warning Year: 2016 Pro-active maize, working hours on drought Period: 2 weeks of harbor terminals Credit rating- Year: 2016 Prioritized policies & Treasury growth strategy Period: 6 weeks investments Year: 2016 Prioritized policies & Phakisa DPME & DAFF Period: 4 weeks investments Year: 2017 Impact of EU imports Chicken crises SAPA & DTI Period: 2 weeks on local industry Range of policies Section 7 Year: 2017 Ministerial including e.g. Crop drought Period: 6 months appointment insurance, finance Impact of WC Year: 2017 DoA WC & pro active Mitigation strategies drought Period: 6 weeks Year: 2017 Mitigation plan to Sugar tax SASA Period: 4 weeks reduce impact Impact of Year: 2018 Impact on growth & SAPA Avian Influenza Period: 3 months jobs, vaccination Land Reform Year: 2018 Pro-active & agbiz N/A N/A Scenarios Period: Ongoing Source: Presentation by Ferdi Meyer, Managing Director, BFAP, 56th l Implemented Agricultural Economics Association of South Africa (AEASA) Conference, l Adopted 25-27 September, 2018 l Pending l No impact 56 What worked less well: Maintaining a modest size has avoided boom and bust cycles of funding and has contributed to maintenance of quality, but has reduced visibility, which may be limiting policy impact. BFAP shares the vulnerability of other African think tanks to funding constraints. Without core funding, maintaining a steady flow of products such as the BFAP outlook reports can be difficult. Lessons learned: The think tank model is especially well suited for environments in which requisite skills already are available, but are not yet organized in a way that engages well with the policy process. The objectivity of think tanks may paradoxically limit their appeal to political actors. In highly partisan environments, politicians may prefer experts clearly willing to provide analysis supportive of favored positions. Finally, a steady flow of scheduled products such as outlook reports and participation in public-private sector forums keep policymakers aware of the existence of BFAP and its ability to respond quickly to queries. 5.4 u Externally-led extended term agricultural policy analysis units in-country: the country programs of IFPRI The operational model: IFPRI undertakes policy-oriented research related to food systems, agricultural growth, rural transformation, and nutrition and has capacity building as one of its core mandates (Kuyvenhoven, 2014). IFPRI’s work on specific countries has often been organized in Country Programs (CP’s) based in the field and implemented in collaboration with local partners. Since 2003 the programs have been designated Country Strategy Support Programs (CSSP’s). CSSPs in Ethiopia, Ghana, Malawi, Mozambique, Nigeria, Democratic Republic of the Congo (DRC), and Uganda were recently assessed as part of a joint exercise by IFPRI’s Impact Assessment Unit (IA) and the monitoring and evaluation unit of the CGIAR Research Program on Policies, Institutions, and Markets (PIM) (Hazell et al., 2018). Operational arrangements of CSSP programs vary. Some are placed within MoAs, and others operate from independent offices, CGIAR campuses, or local research organizations. Most report to one division of IFPRI, typically Development Strategy and Governance. The head of the CSSP is a senior, internationally recruited staff member decentralized from Washington. Other staff include a mix of locally and internationally recruited colleagues, all of whom work according to IFPRI staff rules. Features of the individual programs follow. 57 The Ethiopia Strategy Support Program (ESSP) was launched in 2004. It is based on the Addis Ababa campus of the International Livestock Research Institute (ILRI) and operates as a collaborative program between IFPRI and the Ethiopian Development Research Institute (EDRI). A National Advisory Committee headed by the economic advisor to the Prime Minister and representing key agencies facilitates establishment of research priorities. Capacity building includes joint research with EDRI and close collaboration with the Central Statistical Authority (CSA), as well as workshops and short courses. The Ghana Strategy Support Program is based in an independent office in Accra with an annual budget of about US$3 million. The program has a director and two to three additional international IFPRI staff members plus locally recruited colleagues. The GSSP does not have a national advisory committee, and research priorities are agreed through workshops and seminars convened for planning purposes. The GSSP works with a wide array of partners including government agencies, research organizations, and universities. The Nigeria Strategy Support Program was established in 2007 and is housed in the IFPRI office in Abuja. The program’s main partner is the Federal Ministry of Agriculture and Rural Development (FMARD), and staff work with a wide array of other agencies and organizations. FMARD lacks a functional policy analysis or technical unit, and this makes engagement with policymakers challenging. Collaboration with other agencies and partners partially offsets challenges inherent in work with FMARD. The current Malawi Strategy Support Program (MaSSP) was launched in 2008 with funding from USAID. It is located in a private office in Lilongwe. It has no national advisory committee, and priorities for its annual work plans are agreed between MaSSP staff and USAID. Part of the MaSSP’s USAID funding also involves collaborative research with MSU and the University of Pretoria. Programs have also been active in Mozambique, Uganda, and the Democratic Republic of Congo on a smaller scale and for shorter duration than for the four noted above. The approach to targeting of evidence in policymaking: The IFPRI CSSPs are designed to be responsive to requests for evidence by the key governmental counterparts. These requests are heard either in annual discussions of work plans or 58 on an as needed or fire-fighting basis during the year. Analytical findings are often conveyed personally by the CSSP senior staff to the principals of the counterpart organization (ministry, agency, executive office) that requested the work. The model is thus highly personalized and reliant on informing key personnel who have agency in the policy process. This approach works best in highly centralized political environments (such as Ethiopia) and less well where decision power is organizationally more diffuse (such as Ghana). The recent development of the Kaleidoscope Model of policy influence offers a framework for more broadly- based targeting of evidence, including through use of the media(Resnick et al, 2018). CSSP’s have begun to incorporate elements of this approach. What worked well: The CSSP’s were recently assessed in a mid-term review undertaken jointly by the monitoring and evaluation officer of the CGIAR Research Program on Policies, Institutions, and Markets and IFPRI’s external impact assessment coordinator (Hazell et al, 2018). Individual programs and lines of research within them have also been separately evaluated (DFID, 2013; Renkow and Slade, 2013; Sands, 2008). The new report found that the CSSPs have been appreciated by the host governments and local stakeholders and have generally been assessed positively by donors. A number of individual instances in which CSSPs provided evidence to inform policy change were examined to identify factors leading to success or failure. The reputational capital of the program is found to be very important. The key here is the perception that research and related recommendations are of high quality and objective. Work drawing on models, methods, or data not otherwise available or known to the national counterparts enhances the perceived value of the evidence by offering findings perceived as both new and rigorous. Assessments of policy options based on household surveys and on computable general equilibrium models drawing on updated social accounting matrices have high credibility. National decision-makers particularly appreciate documented lessons from experience elsewhere when these are perceived to be without ideological intent. IFPRI’s ability to augment in-country expertise and bring in international experience is therefore valued. Willingness to adjust work programs to respond to urgent short-term requests is appreciated. Programs with a long track record in place are more likely to have impact than those that are newer or have intermittent in-country presence. 59 What worked less well: The CSSPs are fully donor funded. Host governments sometimes provide office space for units embedded in government offices, but not financial support. This can create issues related to ownership, sustainability, and government perception of undue influence by donors. When funding ends, programs cease without an explicit exit or handover strategy. A single-minded focus on demand-driven analysis can also lead to short- termism, or avoidance of important, but politically sensitive topics. This adds to the dependence of the strategic direction of the program on the wisdom of the steering or advisory committee and the personal skills of the program leader. Staff operate under IFPRI appraisal and promotion rules that emphasize publications and can reduce enthusiasm of junior staff to spend time interacting with national partners or in the capacity development of partners. The CSSP’s have not in all cases identified the best national partners, and this can limit the quality of research, transfer of findings to the right policymakers, or both. Finally, in environments with very active public discourse such as Ghana, the number of initiatives and volume of studies competing for attention is large. Individual studies, even of very high quality, may be difficult for clients and counterparts to absorb. A standardized set of predictable and relevant products (such as Outlook Studies, Public Expenditure Reviews, State of Agriculture, etc.) may distinguish the research program and achieve stronger traction in the policy process. Lessons learned: The CSSP model has had a number of successes. To the extent that quantitative analysis of its impact can be undertaken, the return on investment is acceptable (Hazell et al 2018). Uncertainty about the duration and amount of funding for the CSSPs affects programming and staffing. Meaningful cost sharing by the governments would reduce vulnerability and increase ownership. The CSSPs need to respond to short-term requests, but they also need to undertake longer term analysis and develop data sources that may not have known applications in the short term. Design of annual work programs explicitly around the policy cycle and known measures that will come up for debate and/or decision would likely increase the application of evidence to those decisions. For example, annual budget allocations and periodic design of medium-term expenditure frameworks are 60 predictable elements of the policy cycle, and governments would benefit from regular agricultural public expenditure reviews to guide decision making. The approach to annual work programming with a fully open menu may miss some high-impact opportunities. Joint development of statistical sources and capacity development for central statistical organizations is effective and has long term impact. However, many of the workshops and training sessions on specific topics inform individuals without much impact on institutional capacity more generally. 5.5 u Network strengthening approaches for MoAs where domestic analysis capacity is widespread: the PAPA Project 19 in Senegal’s MoA The operational model: The Projet d’Appui aux Politiques Agricoles (PAPA) was set-up in 2015 by the Ministry of Agriculture and Animal Resources (MARE) of Senegal with USAID financial support and technical assistance from IFPRI and MSU. The purpose was to strengthen Senegal’s policy and enabling environment for increased public and private agriculture sector investment. PAPA was created in response to the realization that while a variety of think-tanks, civil society organizations, and even government-funded research institutes had emerged over the years in Senegal and were working on rural development and agricultural policy issues, MARE decision-makers were not involved in setting the analytical agenda and they could rarely use the output of these organizations when needed to inform their own policy decisions. The technical approach of PAPA features three elements: (a) a network of local centers of expertise empowered financially and administratively to generate knowledge to meet the evidence needs of MARE and other government policymakers; (b) an inclusive policy dialogue and consultation platform allowing a wide group of state and non-state actors to review and influence agricultural policies; and (c) a knowledge management infrastructure that is the portal for disseminating data and analyses to all stakeholders. The approach to targeting of evidence in policymaking: The administrative leadership of PAPA is a full-time function carried out by a MARE executive attached to the office of the Permanent Secretary. Project activities fall under four components: (a) enhancing national capacity for policy research, analysis, and policy communication; (b) fostering political buy-in, stakeholder involvement, and ownership of agricultural policies and processes; (c) promoting evidence- 61 based agricultural policy formulation, implementation, and monitoring and evaluation (M&E), and (d) facilitating effective policy implementation and M&E. The local analytical network facilitated by PAPA since 2015 provides new knowledge on value chains in strategic sectors such as dry cereals (millet, sorghum, maize), irrigated rice, horticultural crops (onion, potato, tomato, banana), agricultural inputs, and equipment. The network has built close to twenty data bases on most issues of concern to MARE. These databases are used to provide reliable and consistent information to public policymakers, the private sector, NGOs, and producer organizations. The network’s research output evolved from producing quick summaries of the findings in the form of factsheets” targeted to policymakers on select topics identified by them to developing a full monograph covering close to twenty topics of interest to agricultural policy-making. To achieve these research outputs, the network organized itself, along with IFPRI and MSU collaborators, into 14 separate groups that worked on analysis in 15 different thematic areas. The work is targeted to providing documented policy options for MARE. Capacity strengthening for MARE central and decentralized units thus far includes training of 196 staff participants, including 81 women receiving training for data collection and analysis, and economic analysis and modeling tools targeting the younger research-oriented staff. PAPA has also provided training to members of producer associations of the inclusive dialog platform that PAPA helped make functional. PAPA strengthened the practice of mutual accountability in policy making by facilitating the Agriculture Joint Sector Revue (JSR) process in Senegal.20 PAPA has also assisted MARE in the production of the policy documents of both the second phase of the Program d’Accélération de la Cadence de l’Agriculture Sénégalaise (PRACAS 2018-2022) and the second generation of the National Investment for Food Security and Nutrition (PNIASAN 2018-2022). During FY2018, an interactive decentralized information technology knowledge management platform and M&E system emerged with PAPA assistance as a locally crafted tool for knowledge management and M&E of the actions of MARE. What worked well: PAPA’s achievements are the result of three factors. First, ownership of the project by MARE top decision-makers has been strong. Second, a consultative approach was used by project leadership in dealing with technical partners broadly defined. Third, the local network of analytical entities 62 has taken increased ownership of research activities over time. After only three years in operation, the sense of ownership by Senegalese institutions is high. What worked less well: With USAID funding coming to an end at the time of writing, it is not yet clear which elements of PAPA will be retained by either the national budget through MARE or other funding. It is also not clear who will play the mentoring role provided thus far by IFPRI and MSU. Thus, a workable model in terms of national ownership and monitorable results may continue to have issues with sustainability once foreign funding is withdrawn. Lessons learned: The PAPA experience and modalities offer useful lessons for African countries struggling to meet the data and analytical needs of more effective agricultural sector policies. The organization of data collection around key data clusters, the coordinated design of data survey and analysis, and the decentralized execution of data and analytical tasks by individual centers of knowledge not only allowed for comprehensive coverage and data on thematic issues, but also was cost efficient. Yet the willingness of MARE to pay for coordination and funding of network knowledge services out of domestic budget remains an unresolved issue. 5.6 u On-call technical assistance from multilaterals having centralized funding: the Monitoring and Analyzing Food 21 and Agricultural Policies (MAFAP/FAO) The operational model: The Monitoring and Analyzing Food and Agricultural Policies (MAFAP) Program is a policy support initiative launched in 2009 by FAO and funded by the Bill and Melinda Gates Foundation, the Netherlands, Germany, and USAID. The program has been implemented in two phases. Phase 1 (2009—2014, with a budget of US$5 million) operated in 10 countries and focused on the development of national capacities to monitor the impact of countries’ policy frameworks on agricultural value chains through analyses of price incentives and public expenditures ex-post. The program’s second phase (2014—2020, with a budget of US$18 million) operates in 13 countries and has expanded the mandate to include closer partnerships with governments on identifying, articulating, and assessing ex-ante policy options. The activities of the MAFAP program work toward the adoption of policy reforms and the strengthening of governments’ commitments and capacities 63 to monitor and analyze sectoral performance. To date, by its own count, MAFAP has influenced over 15 policy changes, mostly in the areas of trade policy, storage policy, strategy, including National Agricultural Investment Plan (NAIP) development, inputs, output marketing, and price policies in agricultural value chains. MAFAP’s team of approximately 15 policy analysts work out of FAO headquarters and regional offices in Dakar and Addis Ababa. At country-level, each partner government has assigned a MAFAP focal point, usually an officer working in the department of policy or planning in the MoA. In addition, FAO has concluded letters of agreement with national research institutes to involve them in policy analysis activities. The approach to targeting evidence in policymaking: MAFAP works on the basis of requests from national governments for analysis and advice on specific agricultural policy issues. After a request, its Rome-based analysts deploy a variety of economic tools to assess the costs and benefits of different options and provide recommendations such as changes in legislation, regulation, strategies, budgets or institutions on the policy reform to be adopted. When needed, MAFAP involves local consultants or researchers in analytical work. In addition, the program collects indicators of agricultural public expenditure and price incentives for all countries in which it operates. This monitoring is carried out in partnership with national MoAs and national research institutes. The program follows a twin-track approach to increase the use of evidence in policymaking. First, it supports policy change based on robust analyses of policy options. Second, it strengthens in-country capacity to monitor the impact of policy on the agricultural sector. What worked well: An independent external evaluator carried out a Mid-Term Review of the program in 2017 (Birner, 2018). A number of strengths, weaknesses, and lessons were identified from this review and follow-up interviews with stakeholders. In terms of what went well, MAFAP managed to achieve a significant number of policy reforms across various areas and countries. The program managed to secure national ownership and buy-in by involving national policy analysts. The evaluation concluded that MAFAP scored very well on providing unique policy advice through: (a) strengthening evidence- based policy dialogue, and (b) articulating policy reform opportunities in terms of content, results, methods and approaches. What worked not so well: The adoption by national governments of MAFAP’s 64 system for policy monitoring has been less successful. Although significant investments were made in technical training for national analysts, these capacities often disappeared during staff rotation or did not reach a sufficient level to function autonomously without FAO support. In addition, results of policy analyses often did not reach leadership levels in the MoA or Cabinet of Ministers as policy staff in MoA lack capacity to coordinate and communicate policy recommendations. FAO’s ability to influence policymaking in ministries other than agriculture is also limited. Lessons learned: First, the capacity to engage across ministries is essential to ensure that related reforms also are considered in policy areas that fall outside the mandate of the MoA. Second, soft skills are extremely important to engage successfully with policymakers. Most ongoing policy support initiatives mainly address technical skill gaps. Soft skills may be just as, if not more critical to strengthening evidence-based policymaking. This is even more the case in policy areas that are multi-sectoral and that require close interaction with other government institutions such as trade, food security, or nutrition policy. The external Mid-Term Review of the MAFAP program also recommended that FAO invest in building and retaining expertise in the management of policy processes as this type of knowledge is essential for providing policy support to national governments. Third, it is important to ensure that issues, tools, and methodologies are nationally owned. The involvement of national policy analysts, either from government or from national think-tanks and research institutes, is a key driver of success in getting results into the policy process. This is illustrated by the ownership demonstrated for MAFAP results by the MoAs in Senegal and Burkina Faso. Furthermore, the degree to which governments are able to use evidence for better policies and public investments is largely determined by the capacities of a small number of policy staff. It is therefore essential to establish close working relationships with key individuals and to invest in their capacities. Fourth, there is a need to advocate for commitment to agriculture beyond the MoA. The fact that policy capacities of MoAs remain weak in many countries in Africa reflects the level of political commitment to agricultural development in other government ministries (especially ministries of finance, planning, development, and trade) and in the offices of the Prime Minister and President. By strengthening the policy capacities of MoAs, institutions will be better able 65 to advocate for policy and investment decisions that are outside of the MoA mandate, but that favor the agricultural sector. Fifth, MAFAP experience also suggests that capacity building funded by donors needs to be better prioritized to fill key gaps. Too often, donor-supported capacity development programs lack clear prioritization of the key capacity bottlenecks in policy analysis, formulation, and implementation. It is therefore essential that support to policy capacity is aligned to a specific country’s institutional context. The Mid-Term Review recommended that assessments of the political economy landscape are consistently carried out before any support is provided. Finally, investing in clear, powerful tools and methodologies pays off. Methodology clearly laid down in handbooks and training materials greatly enhances the buy-in of staff in MoAs. It also increases staff ability to apply them and use them in their day-to-day work even when the external MAFAP team has left. Use of common methods facilitates cross-country comparisons, but may need to be supplemented with approaches specific to the national context. 5.7 u High-level embedded foreign agricultural policy advisors in MoAs: the EU/FAO FIRST Policy 22 Assistance Facility The operational model: Food and Nutrition Security Impact, Sustainability and Transformation Policy Assistance Facility (FIRST) is a large policy assistance program developed and implemented under a partnership between the European Union and FAO since 2016. It currently operates in 24 countries, of which 12 are in Africa. Its total budget for the 2016-2022 period is EUR€41 million. FIRST operates by placing an internationally-recruited, senior policy officer as an embedded policy advisor in the MoA or Ministry of Food Security in each of the countries in which it operates. These advisors assist in: the development and implementation of sectoral and cross-sectoral policy and program frameworks and related instruments; strengthening human and organizational capacities for policy and institutional change; and supporting inclusive and evidence-based governance and stakeholder coordination across sectors and policy domains. The approach to targeting evidence in policymaking: FIRST is highly demand- driven and operates only in countries that responded positively to a call for 66 expressions of interest. Based on the responses to the call, countries were selected. Despite the demand-driven approach, placements primarily have involved countries where the EU has already made or is planning to make significant investments in food security and nutrition projects or provides budget support. The policy officers recruited to be placed in-country are selected based on a specific technical profile requested by countries, which helps improve national ownership. A work plan is agreed upon and signed between the host government, FAO and the Delegation of the EU in each country. This plan sets out the strategic policy areas to be addressed by the policy officer. Typically, these build on government and donor priorities and span areas such as agriculture, food security, nutrition, livestock, or land use. Policy officers manage a FIRST-provided budget of US$100,000/year to organize meetings or recruit consultants in support of their evidence-based policy work. Between 2016 and 2018, the focus of the officers working on the program was mainly on policy development. Over 30 policy reforms were adopted that can be mapped back to advisor involvement. A national budget was allocated to these reforms in most cases. Since 2019, the focus of the program has shifted towards implementation capacity. This is to address the key constraints that hold back government capacity to deliver on the adopted policy instruments, investment plans, and EU budget support programs. What worked well: A key success factor for the FIRST program was its close partnership between government, a technical organization (FAO), and an influential donor (EU). As much as possible, FIRST worked in countries where the EU had already made or was planning to make significant investments in food security and nutrition projects, or where it provided budget support. FIRST could add value to those investments. Second, FIRST was successful in striking a balance between opportunistic, ad-hoc, or short-run policy advice on-demand and delivering on a strategic agenda. This strategic agenda was facilitated by negotiating a clear inception plan signed by the host government, FAO, and the EU at the start of each assignment. Third, FIRST was successful in building a network of policy officers with a wide variety of skills. This allowed the program to respond to policy requests in a variety of technical areas across countries, despite the one-person-per-country model. What worked not so well: First, the timeline used by FIRST was too tight. 67 Originally, the period of engagement for each policy officer was two years. The experience of FIRST shows that this was too short to deliver results. Second, the program faced major challenges in finding well-qualified policy officers. The profile of the individual officers determines success or failure in influencing policy. Therefore, the right mix of technical and soft skills is essential. However, persons that combine these two are hard to find. Lessons learned: Given its scale and scope, the first three years of the program offer important lessons. First, a sufficiently long time horizon is required for policy support initiatives to achieve results. Second, sharing of expertise across countries when specific need arises is important, especially in the case of a single person per country model. Third, policy officers need soft skills essential to understanding the political economy, context, and institutional power relations of the host country. They must also be able to handle political sensitivities successfully. Fourth, successful policy support programs require staff with an understanding of the budget process and cycle of the host country to ensure that budgets are allocated to and spent as agreed for implementation of the policies, plans, and strategies. Fifth, big project infrastructure and branding should be avoided to temper resentment and build trust. Modesty and a low profile also allow policy officers to participate more actively in donor working groups, join sector reviews, and other existing structures of policy dialogue without being considered a threat. Recruitment of the right persons is essential but difficult. It is a constraint to scalability of the model. 5.8 u High-level embedded foreign agricultural policy advisors in the Executive: the Tony Blair Institute for 23 Global Change (TBI) The operational model: Started in 2008 as the Africa Governance Institute, the TBI provides support to governments in Africa to strengthen management and the delivery of government policy in a variety of areas including agriculture. It operates in 14 countries across Africa, mostly in West Africa. Following a government request, TBI supports governments in four key areas: (a) high-level political advice delivered by Tony Blair personally to African leaders; (b) capacity-building on coordination systems and best practices in public management; (c) delivery support to help governments drive the implementation of policies and programs, and (d) strategy and policy analysis 68 such as the preparation of policy or implementation briefs. TBI’s programs are led by in-country delivery advisors that work as a team spanning various ministries and agencies. It is often anchored in the Office of the President to help provide support to the development agenda of the center of government. Besides agriculture, TBI works on sectors such as infrastructure, manufacturing, education, and health. The size of in-country teams differs from a few persons to a group of 18 staff in Ethiopia, where TBI has its biggest program. The approach to targeting evidence in policymaking: TBI’s agriculture policy support predominantly addresses the linkages between agriculture and other sectors through a network of advisors embedded in government institutions. In Liberia, TBI supported coordination and alignment across the Ministry of Agriculture, Ministry of Trade and Industry, National Investment Commission, Ministry of Finance and the Presidency. In Sierra Leone, TBI supported the coordination and delivery of agricultural policies with the Ministry of Finance and the President. In Rwanda, TBI has advisors in the National Agricultural Export Board (NAEB). Other countries where TBI works (or is soon starting to work) include Cote d’Ivoire, Ghana, Mozambique, Burkina Faso, Kenya, Togo, and Nigeria. What works well: The model of embedded advisors operating on a broader, multi-sectoral in-country team that provides management and delivery support to the highest levels of government ensures that TBI support is aligned with executive branch ownership. Through direct contacts between the TBI Chairman and Heads of State, TBI is able to support the emergence of a coordinated approach to agriculture transformation across the key ministries and agencies. It does this while using the political agenda and an understanding of the political economy to drive systemic change. As a result, TBI’s support benefits from a high degree of country ownership. The TBI model also allows the institute to work across ministry lines and to elevate agricultural transformation from being a MoA issue to a topic at the center of government. What works not so well: The institute works internally with relatively small teams that typically have not been resourced for policy planning, analysis, monitoring, and evaluation capacity in MoAs and other institutions responsible for policy reform that affects the agricultural sector, such as trade and industry. In addition, TBI does not carry out data collection, nor seek to execute in-depth primary research. Where research views are sought from TBI by the government 69 served, the focus is on synthesizing existing research into forms relevant and manageable for policy makers. Lessons learned: TBI’s experience shows that building an in-depth understanding of the political economy context is essential for policy support. MoAs are often characterized by patronage systems and weak implementation capacity that severely constrain their abilities to support market systems in agriculture. Yet it is essential to work with private sector representatives to ensure that policy processes are inclusive and address the needs of all agricultural sector stakeholders. Support from other government ministries (especially trade, investment, and finance) and the President’s office greatly facilitates rapid change in agriculture policy, at least at the policy adoption stage. 5.9 u Government agricultural policy and monitoring units outside MoA: the Ethiopian Agricultural 24 Transformation Agency The operational model: The Ethiopian Agricultural Transformation Agency (ATA), a government agency reporting to the Prime Minister’s Office, focuses on strategy and delivery to accelerate the growth and transformation of Ethiopia’s agricultural sector. It was established in 2010. The ATA is widely considered as an institutional approach that has successfully furthered agricultural policy implementation, and through this, has served as a key factor in the rapid development of the country’s agricultural sector in the past decade. Currently, the ATA is halfway through its organizational lifespan, which ends in 2030. The approach to targeting of evidence in policymaking: ATA’s role is threefold. First, it seeks to identify the bottlenecks within the Government of Ethiopia Agricultural Transformation Agenda. Second, the ATA supports the integration of solutions to address crop commodity bottlenecks through the Agricultural Commercialization Clusters Initiative (ACC), which is implemented through regional governments. And finally, the ATA supports the Ministry of Agriculture and Livestock Resources (MoALR) on the development of livestock commodity value chains. Implementation of ATA strategies falls into the mandate of other government institutions, such as MoALR. These institutions suffer from weak capacity and 70 limited financial and human resources. As a result, not all strategies developed by the ATA can be implemented in a timely fashion. In order to remove some of the initial risks and barriers, the ATA engages in policy implementation by undertaking dedicated pilots, before handing them over to other government institutions for wider scale-up. This approach was followed in a number of key areas, including: • The input voucher system: a new input subsidy voucher system was piloted in one region and less than 10 districts by the ATA itself before roll-out nation-wide. • ICT for agricultural services: ATA piloted the 8028 Farmer Hotline, a service that provides advice on agronomic best practices directly to farmers through mobile phones and landlines. • Mechanization: ATA has started a new pilot on mechanization that focuses on the establishment of service centers providing farmers with access to tractors, combine harvesters and other equipment. What worked well: A key area of strength of the ATA model is its development of advanced analytical capacity to carry out its mandate. This analytical capacity has been built in-house in a dedicated analytics unit modeled after large international consultancy firms. During the few years of ATA’s existence from 2011 to 2015, it focused on the preparation of strategy documents on how to address bottlenecks in various sectors, such as seed, mechanization, rural finance, etc. The strategies presented clear recommendations classified by time horizon (short term—long term) and type of intervention (role of government, institutions involved, policy instruments to be used, etc.). What worked less well: The relationship between the ATA and the MoALR was a big challenge as ministry staff felt threatened by a new agency that intruded on the ministry’s mandate. Now, several years on, this has changed. ATA and MoALR reportedly enjoy a normal working relationship and the MoALR channels a large number of questions and requests to the ATA. The driving factor of that was that the ATA not only assumed responsibilities that were originally within MoALR, but also gave back through a delivery unit within MoALR and 20 full-time ATA staff assigned to the MoALR. They support the tracking and reporting on the transformation agenda and provide technical 71 assistance to policy implementation. By doing this, the ATA has contributed directly to the strengthening of the MoALR’s institutional capacity. Lessons learned: Three key lessons emerge. First, a clear mandate and high- level political commitment to agricultural transformation are essential to ensure that other government institutions, and especially ministries of finance, trade, infrastructure, and energy, consider agriculture a priority. In the case of the ATA, several ministries initially viewed the ATA unfavorably. However, this gradually faded over time. Demonstrating excellence in delivery was of great importance for the ATA to be taken seriously. Second, as illustrated above, empowerment of a super implementation unit can be effective, but resentments need active management. Third, the ATA is a model idiosyncratic to Ethiopia that benefited from high-level political support and commitment, strong government powers, and long-term external financial and technical support for agriculture. ATA also operates under the salary and human resource rules outside the constraints of the civil service system, and hence has flexibility in staffing not within reach of line ministries. 5.10 u An African-led regional policy analysis network to strengthen CAADP: ReSAKSS The operational model: The Regional Strategic Analysis and Knowledge Support System (ReSAKSS) was established in 2006 to promote evidence-based policy planning and implementation in support of CAADP. ReSAKSS provides data, analysis, and related knowledge products to facilitate CAADP benchmarking, policy dialogue, review, and mutual learning processes. It was established by IFPRI in partnership with AUC, the African Union Development Agency (AUDA-NEPAD), three regional economic communities (RECs)–ECOWAS, COMESA and SADC, and Africa-based CGIAR centers. ReSAKSS operates as a knowledge-brokering network represented by four platforms: one at the continental or Africa-wide level (ReSAKSS-AW) and three at the sub-regional level—East and Central Africa (ReSAKSS-ECA), Southern Africa (ReSAKSS-SA), and West Africa (ReSAKSS-WA). The sub-regional nodes are hosted by the International Livestock Research Institute in Kenya, the International Water Management Institute in South Africa, and the International Institute of Tropical Agriculture in Nigeria, respectively. Each sub-regional node is governed by a steering committee consisting of representatives from major 72 CAADP stakeholder groups and chaired by the respective RECs: ECOWAS for ReSAKSS-WA, COMESA for ReSAKSS-ECA, and SADC for ReSAKSS-SA. ReSAKSS-AW activities are coordinated by IFPRI under a governance structure that is chaired by AUC and AUDA-NEPAD. The steering committees provide oversight that ReSAKSS activities remain focused on providing policy-relevant data and analysis in support of CAADP. ReSAKSS stakeholders and partners include government agencies, policymakers, policy analysts, research institutes and researchers, the academic community, civil society, farmers’ groups, the private sector, and donors. ReSAKSS also operates at the country level through Strategic Analysis and Knowledge Support Systems (SAKSS) in 13 countries, platforms that work through local centers of expertise to support the review of and dialogue on CAADP and provide country-level data and analyses. ReSAKSS provides data, analytical tools, and technical support to: (a) monitor CAADP progress using interactive online platforms and a flagship Annual Trends and Outlook Report (ATOR); (b) conduct strategic analysis through the ATOR and occasional studies; (c) promote evidence-based national agriculture investment plans (NAIPs); (d) strengthen capacities for policy analysis through training workshops and working collaboratively with in-country and regional experts; and (e) enhance mutual accountability processes such as JSRs at the national and regional levels and through the continental CAADP BR. The latter contributes to improving policy planning and implementation processes, a key feature of the CAADP agenda. ReSAKSS carries out research and policy analysis. These are limited to the CAADP M&E, the annual flagship publication, and occasional working papers. The bulk of ReSAKSS work is carried out in collaboration with others and in- depth national assessments are contracted out to qualified local experts. ReSAKSS supported 47 countries to implement the inaugural CAADP BR in 2017. The approach to targeting of evidence in policymaking: ReSAKSS responds directly to the knowledge demands emanating from the CAADP agenda through requests from AUC and AUDA-NEPAD, RECs, and national governments. ReSAKSS has been regularly called upon by AUC, AUDA-NEPAD, and national governments to provide country teams with analytical support to guide the formulation of second-generation national agriculture investment 73 plans (NAIPs). ReSAKSS accordingly has: (a) developed an AUC endorsed NAIP Toolbox that outlines metrics, analytical tools, methodologies, and core programmatic components to guide NAIP design; (b) set up a NAIP Task Force made up of international researchers to provide training, backstopping, and quality control; and (c) mobilized and trained over 200 local research experts to deliver analytical support to nearly 30 countries. Most of the trained local experts are members of the African Growth and Development Policy Modeling Consortium (AGRODEP), which was set up by IFPRI in 2010 as a knowledge generator to complement ReSAKSS. Policy issues covered in the ATOR and occasional papers are identified through consultations with members of the ReSAKSS steering committees. The 2009 ATOR focused on the 2007/2008 global food price crisis and considered the growing intensity and frequency of climate change effects. The 2016 ATOR examined the role of climate-smart agriculture in helping to meet the Malabo Declaration commitment to enhance the resilience of livelihoods and production systems. In addition, ReSAKSS widely disseminates CAADP data and analytical knowledge products through its websites and workshops, meetings, and conferences held at the national, sub-regional, and continental levels. What worked well: Conducted at the request of USAID, a 2015 independent performance evaluation of ReSAKSS by Social Impact Inc highlighted several areas in which ReSAKSS is performing well. (a) Timely and Policy Relevant Support: According to the evaluation, ReSAKSS is providing the right kind and level of information to support CAADP processes. It noted that ReSAKSS served as a bridge that links country-level analysis to regional and continental policy issues. ReSAKSS stakeholders found the data syntheses produced by ReSAKSS through the ATOR to be particularly useful in informing their own planning and analytical needs. ReSAKSS has been able to effectively respond to CAADP knowledge demand by working closely with key CAADP institutions. AUC, AUDA-NEPAD, RECs, and country governments regularly call on ReSAKSS for technical and analytical support. (b) Contribution to Use of Policy Analysis: During the first phase of its support to CAADP from 2006 to 2009, ReSAKSS played a central role in helping to transition CAADP from a program document to field implementation. In partnership with IFPRI, ReSAKSS conducted analyses that defined strategic agricultural priorities 74 and funding options to increase growth and reduce poverty. The analysis fed into CAADP roundtable meetings and informed the development of first- generation NAIPs in roughly 30 countries. More recently, ReSAKSS partnered with local experts to inform the formulation of second-generation NAIPs in 30 countries, including all 15 ECOWAS member states. Several of the countries acknowledged the analytical support of IFPRI/ReSAKSS in their NAIPs. Examples include Burkina Faso, Cote d’Ivoire, Guinea, Mali, and Togo. Other second generation NAIPs, such as for Cote d’Ivoire (2017-2025) and Senegal (2018-2022) extensively reference IFPRI/ReSAKSS analysis, which is used to identify priority agricultural commodity value chains for targeting investments. (c) Strengthened Analytical Capacities: The ReSAKSS model has helped to strengthen local capacities for data collection, policy analysis, economic modeling, M&E, and policy review through the JSR and BR processes. The 2015 performance evaluation of ReSAKSS highlighted the important role ReSAKSS has played in strengthening capacity for agricultural policymaking through training and collaborative work. Through the partnership with the African Growth and Development Policy Modeling Consortium (AGRODEP), ReSAKSS strengthened economic modeling capacities of over 200 local experts to conduct analyses to guide NAIP prioritization and investments.25 These experts are now being deployed to help inform country JSRs, the continental BR, and the preparation of NAIPs. (d) Improved Policy Review and Dialogue Processes: ReSAKSS has contributed to improving the quality of agricultural mutual accountability processes. Since 2014, ReSAKSS has conducted assessments of JSRs in 23 countries and two RECs (ECOWAS and EAC) to evaluate the quality of the current review processes and ways to improve them. Today, JSRs have become more inclusive of non- state actors, more comprehensive in coverage, and have better monitoring and follow-up of actions. This has led to improvements in policy review in countries such as Malawi and helped launch new JSR processes in Burkina Faso and Senegal. The evaluation by Social Impact noted the important contribution ReSAKSS has made in helping to standardize the JSR process, to report on outcomes, and to encourage country governments to bring in more non-state actors to participate in the process. What worked less well: The 2015 Social Impact evaluation also noted some inconsistencies in ReSAKSS support due to it being overstretched. Despite its 75 small size, ReSAKSS has at times been called upon to support the BR process in all participating African Union members states, something that is difficult to do. The evaluation also highlighted that the information provided by ReSAKSS can be too academic and lengthy. ReSAKSS has since worked to address this by presenting its work in a variety of more readily accessible, concise, and easy to digest formats such as policy briefs, flyers, and brochures. ReSAKSS is strengthening its partnerships with others such as the CAADP Non-State Actor Coalition, which can use succinct products for outreach. Lessons learned: The establishment of direct links to and drawing a mandate from continental and regional inter-governmental organizations creates the political ownership and regional legitimacy for ReSAKSS critical to demand for its product and their applications. This both creates the space required for regional decision-making and encourages demand by member states. As a knowledge broker (as opposed to a knowledge generator), ReSAKSS relies on a wide web of collaborations that have allowed it to operate with a limited number of staff and kept operational costs low relative to its contribution to promoting better use of evidence in agricultural policy decision-making. Low cost and the legitimacy conferred by regional ownership should both add to the long-term sustainability of support for ReSAKSS. 5.11 u An African-led regional policy analysis network built around a university identity: ReNAPRI The operational model: Initiated in 2012 and formally launched in 2014, the Regional Network of Agricultural Policy Research Institutes (ReNAPRI) provides a forum through which members can coordinate activities, share data, and collaborate on various research and policy initiatives of regional importance. It also facilitates the cross-fertilization that derives from an understanding of shared experiences of other countries within the region.26 In ReNAPRI’s case, first MSU’s and then IAPRI’s work in Zambia (see section 5.2 above) on regional trade and seed and fertilizer subsidy policies had previously led to a series of partnerships with regional policy institutes, most of them based at agricultural universities such as Pretoria, Stellenbosch, Sokoine, and Tegemeo. These ongoing research collaborations contributed to the formation of a formal association of regional agricultural policy institutes under ReNAPRI. Like its constituent think thanks, ReNAPRI contributes more to the generation 76 of research for policy analysis than to the consolidation and translation of existing evidence for use by policymakers. But it is still engaged in the latter, as are other policy think tanks (see the case of ReNAPRI members INDABA in Section 5.2 and BFAP in Section 5.3, for example). ReNAPRI’s sub-regional status and ownership suggests considering it as a fledgling companion effort on the research side to regional knowledge-brokers such as ReSAKSS. ReNAPRI was formally registered as a network in Zambia in 2014 with a Secretariat and a Board of Directors drawn from the management of member institutions (Jayne et al. 2019). ReNAPRI includes the following founding members:27 • The Institute of Social and Economic Research (IRES), University of Kinshasa, Democratic Republic of Congo; • The Tegemeo Institute of Agricultural Policy and Development, Egerton University, Kenya; • The Centre for Agriculture Research and Development (CARD), Bunda College, Malawi; • The Sokoine University of Agriculture (SUA), Tanzania; • The Research Center for Agricultural and Food Policies and Programs (CEPPAG), Eduardo Mondlane University, Mozambique; • The Bureau for Food and Agricultural Policy (BFAP), Universities of Pretoria and Stellenbosch, South Africa; • The Indaba Agriculture Policy Research Institute (IAPRI), Zambia; • Department of Agribusiness and Natural Resource Economics, Makerere University, Uganda; and • Department of Agricultural Economics and Extension, University of Zimbabwe. The approach to targeting of evidence in policymaking: Ownership by nationals is critical to success in policy advice in Africa today. Individual ReNAPRI members, especially the stronger ones like IAPRI, have arguably established brands with their national policymakers. Regional brands add to, but do not compete with national brands. ReNAPRI can be understood as a way of increasing the regional brand of the agricultural policy research its members undertake. 77 ReNAPRI has a number of institutional partners, including non-African ones, beyond its core membership. These include: two U.S. universities with long commitment to collaboration and capacity building in the region, MSU and the University of Missouri; two inter-governmental organizations, the United Nations Food and Agricultural Organization and the Common Market for Eastern and Southern Africa (COMESA); and three regional agricultural policy research organizations, ReSAKSS, the Food and Natural Resources Policy Analysis Network (FANRPAN) and Center for Coordination of Agricultural Research and Development of Southern Africa (CCARDESA). Arguably there is a degree of overlap in functions and funding sources between ReNAPRI and its institutional partners, with each institutional partner able to deal directly with each ReNAPRI member institute. Presumably the ReNAPRI structure strengthens the hand of individual member institutes in dealing with institutional partners compared to going it alone. This is both through increased information-sharing and through ReNAPRI having a degree of regional legitimacy on its own. Targeting of specific evidence to policymakers happens through the local member institute of ReNAPRI in the country concerned. What worked well: In a relatively short period of time, ReNAPRI has succeeded in achieving an appreciable degree of regional legitimacy with African inter- governmental organizations such as COMESA, national governments, funders, collaborators, and the rising regional intelligentsia (Jayne et al. 2019). This has occurred in an increasingly crowded field that is jockeying for regional legitimacy and funding. It seems to have strengthened the hand of its members in dealing with external and regional organizations. There also seems to be actual collaboration between members that could lead to policy synergies across borders (Ibid.) ReNAPRI has also extended the influence of its stronger members such as IAPRI (Section 5.2) and BFAP (Section 5.3). What worked less well: A cornerstone of regional organizations in Africa has been the principle of equality of all member countries, large and small. The same applies to agricultural policy research institutions in a network such as ReNAPRI. Yet reality is quite different in terms of staffing, analytical strength, and access to funding across member institutions. In an environment where funders continue to pick and choose, this introduces the potential for discomfort when some members feel that their turf is being encroached upon by other members (Jayne et al. 2019). ReNAPRI, like most regional organizations in 78 Africa, is still working on how to reconcile the one-country-one-vote philosophy of the region with the realities of the marketplace for funding. Lessons learned: ReNAPRI adds value mainly through the contributions of its more established members. That value will only be larger than the sum of its parts at a regional scale if the strengths of the more established partners are systematically drawn upon and assimilated by the newer and weaker members. This is still an open issue and may require proactive involvement by the governments of the countries requiring capacity building services from the stronger members of ReNAPRI. 79 6 80 Synthesis of Lesson-Learning from the Cases The discussion of various models summarized above confirms the range and depth of experience over time in Africa. Efforts to consolidate and deliver evidence relevant to policy decisions are not new. The fact that such a variety persists suggests that no single model may be superior in all aspects for all cases. Historically, there may have been a real or perceived difference in the quality of analysis and recommendations on agricultural policy from teams that had significant external direction, finance, and intellectual input, and were more home grown. In periods of acrimonious policy dialog, such as the difficult period of the early 1980s, this may have led in some cases to a trade-off between quality of policy analysis in terms of evidence-base versus national or regional ownership. The main solutions to this over time were essentially twofold. One approach was to build a national institution like ATA in Ethiopia for implementation or IAPRI in Zambia for analysis, where adequate funding and political support produced a national institution that is truly internationally competitive in generating or plugging into evidence-based analyses and is well-owned by its host country. A different, more regional approach, like ReSAKSS or ReNAPRI, targeted economies of scale and scope in the generation of quality policy evidence, analysis, and delivery while building an Africa-based and owned brand. A trade-off remains between the incentives that generally are most conducive to the generation of good policy evidence and being in a position to be effective deliverers of that evidence to policymakers when needed in forms they can use. The former requires high analytical skills and matching compensation, access to a wide array of funding sources, intellectual freedom, accessible contacts abroad, and an ability to accept private sector incentives. The latter requires trust on the part of decision-makers including seeing their advisors as accountable to their ministry and not some other institution; an understanding of what policymakers actually need to know; an ability to succinctly and dispassionately deliver on time according to the policy calendar across options; 81 and delivery without a need for personal intellectual ownership or recognition for a particular solution. Resolving this trade-off is where solutions to resolving the “weak link” are most likely to be found. The cases examined suggest the following take-aways in finding those solutions. 6.1 u Ownership by national decision-makers and trust matter Stakeholders with agency in the policy process must be confident that they can make their needs known, and that policy analysts will address these needs with professionalism and timeliness. In this sense, the effort must be owned by stakeholders, particularly those with the greatest power in the policy process. This point was emphasized time and again in the companion workshop to this paper (see Annex I and Sakane et. al. 2019) and in our analysis of case studies. African countries are clearly moving on from an era where non-nationals make major decisions in domestic and regional policy domains. Given the importance of ownership, it is curious that so few of the examples reviewed were or are now funded nationally. Most depend wholly on external funding. A model of shared initial funding with transition to long-term national funding through the government budget and local contracts would seem best placed to address the issue of ownership. Oversight by a Board of Trustees or Steering Committee perhaps could safeguard objectivity should ownership exert pressure. The African Development Bank (AfDB) recently stressed two principles for guiding its support to evidence-based agricultural policymaking in Africa going forward.28 First, there will not be new AfDB support for new institutions. Second, AfDB support will be contingent on significant domestic fiscal support being made available. The latter is a clear statement of a key metric for assessing national ownership of evidence-based policy analysis. 6.2 u Longevity matters in the establishment of trust In discussion during the workshop May 1-2, reference was often made to a policy ecosystem as the environment in which policy is made (Sakane, Delgado and Rege, 2019). Systematic embedding of evidence within the policy process requires that the sources of evidence have a home within the ecosystem. 82 Experience summarized in the paragraphs above illustrates the long time- period required for establishment of such a niche or home in the ecosystem, institutional memory and learning, and accrual of reputation underpinning trust. The 15 years of previous collaboration before the founding of IAPRI is one example. The entity undertaking and providing policy advice must have been around long enough to be known and have gained the trust of the political class, members of which may change functions over electoral cycles. The entity must, moreover, have an institutional identity that can withstand changes in personnel and can span funding cycles. Longevity matters particularly if the intent is to have a steady flow of evidence to inform policymaking over an extended timeline. If the intent is instead to seek advice on a specific and time-bound issue, then external technical assistance can be commissioned, although successful sourcing of the work and interpretation of results will likely require domestic capacity. The Tanzanian Ministry of Agriculture’s Department of Policy and Planning is reported to serve largely the latter function through sourcing and oversight of analysis done by outside experts (Isinika, 2016). Given the complexity, importance, and recurring nature of agricultural and food policy issues in Africa, it is difficult to argue that analytical capacity is a short-term need potentially addressed by temporary measures. Institutionalization of analytical capacity to inform policy choices is a feature of the environment in middle and high-income countries. A path to the establishment of organizations that can fulfill similar functions is required in Africa today. Several of the IFPRI Country Programs and the South African BFAP have continuity sufficient to establish relationships of trust with policymakers and civil society organizations and relationships with local research partners. The Ethiopian ATA is also currently firmly established and integrated into the policy ecosystem. Given the importance of longevity, any analytical entity or network should be established with commitments of five years of initial funding from shared sources and an explicit decision point at which a determination is made as to the utility of continuation from identified national funding sources. During the initial five-year period, the principals will be expected to invest in relationships that confirm the initiative’s usefulness and underpin a decision for continued support. 83 6.3 u The record of delivery units within a Ministry of Agriculture has not been impressive Efforts to create policy support functions as units that address the weak link between the availability and use of policy evidence within African MoAs have foundered for several reasons. Civil service pay scales and incentive systems that govern MoAs do not serve to recruit and retain staff with highly specialized skills in demand outside of government. Skills of agricultural policy analysts fall into this category, with high demand in the NGO sector and some demand in the private sector. Training provided to ministerial staff often serves to facilitate exit of the most capable members and concentration of the remaining colleagues in the public sector. Under these circumstances, public service does not carry enough prestige to counter the poor pay. Governments and donors facing these constraints have sometimes resorted to enclaving, moving defined functions out of government or to parastatals to allow different pay scales and promotion opportunities. The challenges of politics and capacity constraints have led some countries to experiment with enclave approaches to civil service reform, spinning off selected government entities from central government ministries (World Bank, 2005). The ATA in Ethiopia was established outside the MoARL in part for this reason and explicitly seeks to align its salary and benefits packages with international counterparts, which is not the case for the Ethiopian civil service.29 The organizational culture within MoARL often stresses politics, process, and practice, rather than the technical and analytical excellence required for good policy analysis. Even if the organizational culture within a MoA is hospitable, the policy process entails other ministries and parliamentary committees. As a result, placement within the MoA may limit exposure of relevant parties to analytical outputs. Although the lessons of experience do not indicate a single best home for an analytical unit, they clearly suggest that MoAs will rarely be contenders. At a minimum, MoAs need internal capacity to understand the policymaking and analytical issues at stake, to manage budgets for analysis of options, to prepare adequate terms of reference for consultants, to monitor externally commissioned work, and to analyze, broker, and communicate results to policymakers. The PAPA case study offers an intriguing model that might work for MoAs in 84 cases where there is sufficient local analytical capacity already built into local think tanks and NGOs. However, PAPA’s viability in the absence of an IFPRI or MSU to support analytical coordination is not clear, nor is the case that MoAs are really prepared to support a significant share of the costs of such an activity. 6.4 u Implementation of agricultural strategy based on evidence requires the ability to work with entities beyond MoAs A neutral base as an independent government agency appears to have an advantage in coordinating the many different government functions necessary to implement an agricultural transformation strategy. An entity with such a base would need to have strong support from the Prime Minister or President’s Office and strong links to relevant ministries, parliamentary committees, media, and trade associations. It would also need to be able to relate to providers of evidence and forecasting, such as think tanks. Finally, it would also need to be deeply familiar with the realities of implementation. The ATA in Ethiopia has such a capacity to work across governmental units or silos and with outside entities. In World Bank terminology, the ATA functions much as a “Mega Project Implementation Unit” within government. Replication of the arrangement in other countries would require having a large public agricultural investment program, funder interest in directly supporting implementation, a strong government presence, coordination with local governments, and exemptions from civil service personnel rules. The experience presented above suggests that having such an institution in government can create frictions with line ministries that end up weakening the latter, especially MoA in the case of implementing agricultural strategy. A high level of managerial competence is needed in an independent government unit to broker relations with other agencies and deliver results. Finally, the sustainability of the independent agency approach is also an issue if the efficiency advantage stems from a waiver of civil service rules and higher pay due to external funding. An end to external funding could result in both a weakened line ministry and closure or serious weakening of the independent agency that had taken over part of the ministry’s implementation role. If an enclave entity is considered a short-term measure to cover acute needs while 85 capacity within the line ministries is enhanced, then the plans for capacity development become important, as does the exit strategy. 6.5 u Placement within a Prime Minister’s or President’s office can undermine sustainability Creation of an agricultural policy support function directly under the chief executive, but without adding a large new institution such as ATA, can facilitate coordination between and among line ministries. It can also assure high-level access to policymakers. However, such placement close to the locus of political power while still relying on line ministries for all oversight of implementation can lead to disconnects between policymaking and policy implementation. It can also raise perceptions of politicization in agricultural policy, thereby undermining trust more widely within the policy ecosystem. Moreover, close affiliation with the head of an administration can undermine longevity by reducing the likelihood that the unit will weather a change in administrations. This is especially likely to be the case if embedded advisors are high profile foreigners. A more neutral base outside government with strong connections to the relevant governmental ministries and offices appears wiser in many cases. Trust, the indispensable requirement for successful absorption of evidence- based policy analysis, may ebb and flow as political administrations change. Work in support of a policy analysis functions must be perceived to be of good quality and objective. Since policymakers will not in general be in a position to evaluate the methodology leading to results, they must have trust in the researchers and particularly in the management of the underlying research effort. Trust allows them to accept the results and recommendations without replicating them. Trust takes time. Placement within executive offices may not allow the required time or contribute to confidence in the objectivity of the advisory function. 6.6 u Independent think tanks are good for policy research, but face challenges in moving research findings into use in the policy process The discussion of independent African agricultural policy research think-tanks above suggests that they are strong models for identifying policy research issues that will be important down the road. The provision of foresight and 86 preparation for future policy needs can be addressed through annual outlook publications that surface issues and invite feedback. An example would be the BFAP outlook reports that inform on questions the policymakers may not yet be asking. Think tanks are also good at fundraising, motivating key national talent, and generating research-based evidence that is free of political bias. They are vital parts of the agricultural policy value chain discussed in the introduction. However, they are typically not well-structured to address the weak link as defined in this paper. Think tanks reliant on external funding from diverse sources are vulnerable to fragmentation of priorities from different donors wanting different things as soon as possible. Further, staff who do well at research are not necessarily the right people to present different options dispassionately for policy decisions at short notice and without recognition of authorship. Another issue is that much of the analytical work by think tanks—especially those with an in-depth research goal—is typically designed without explicit reference to specific policy decisions to be debated according to a known calendar. Government agencies are typically better on timing and other entry points to what has been shown to be a multi-faceted and complex policy process (Resnick et al. 2018). Think tanks need to work actively on building government trust in their relevance, responsiveness, and impartiality. Capacities for impartial analytical vision and relevance to current policy needs can be brought closer together by efforts on the part of government, the private sector, independent analytical institutes, and external donors to coordinate through periodic meetings. Tanzania provides an example of this via the successful Policy Analysis Group and Annual Agricultural Policy Conference, a coordination mechanism for agricultural policy analysis in Tanzania with substantial input from MSU (USAID, 2017). The goal should be to make such coordination less dependent on external funding over time. Over the longer run, two things will likely be necessary for African think tanks to be able to play a bigger role in agricultural policy consolidation and translation sustainably in their own countries. The first is stronger host country government uptake of local analytical entities through co-financing of coordination mechanisms and capacity-building for domestic think tanks and NGOs. Second, more effective donor coordination under government auspices is needed in the funding of policy analysis. This seems a necessary evolution 87 both for governments and agricultural policy think tanks to build a more functional agricultural policy ecosystem. The growing emphasis on regional collaboration in policy analysis and data should help by unifying approaches and providing increased visibility and quality assurance. 6.7 u Embedding foreign advisors is preferable to short- term technical assistance missions but should include specific efforts to build national capacity Using externally funded and sourced embedded advisors in national agricultural policy processes can be useful short-term assistance to addressing the weak link. It is generally preferable in this case to short-term externally sourced and funded technical assistance missions, at least from the standpoint of the host country. However, embedded advisors are most useful when combined with a specific institutional capacity building effort to permit nationalization of the embedded advisors’ functions longer term. The willingness of governments to contribute to the cost of such capacity-building should be factored into the allocation of advisors by funding agencies. 6.8 u Regional networks can help build ownership in defining regional policy options while capturing economies of scale and scope from a regional approach Regional approaches help where countries are too small or problems are too complex to maintain adequate national capacity to address the weak link on their own. Regional collaboration has worked well on the generation (i.e. the supply side) of evidence for agricultural policymaking as in the cases of ReNAPRI. The demand side is more problematic as policy tends to be largely a national matter and distrust across countries common. CAADP is a regional initiative. It advocates for specific regional approaches to being strategic, inclusive, and evidence based. However, the resulting NAIPS are still national in nature. Knowledge- brokering at the regional level, as through ReSAKSS, offers countries access to economies of scale in accessing the knowledge for implementation of national priorities. Government-to-government dialog within regional organizations also suggests the possibility of at least capturing some economies of scale and scope from policy formulation work conducted through regional networks. However, national ministries still need the capacity to participate and advise national policymakers. Keeping MoAs permanently linked to knowledge-brokering 88 networks such as ReSAKSS through regular meetings and other activities is a low cost way to reduce transaction costs when rapid response is needed between national entities and regional databases and researchers. 6.9 u Success should be measured by contribution and influence rather than policy outcomes The motivation for providing evidence should be to inform the key agents in the policy process, rather than to guarantee a particular outcome. The political economy of agricultural policy is complex, and actors have mixed objectives. Success stories in which research products are broadly understood to have contributed to outcomes consistent with growth, poverty reduction, attainment of SDG’s, and other desirable outcomes should be tracked and publicized. The metric of success, however, should be the contribution of the research to informed debate, rather than attribution of an outcome to research products. Contribution can be tracked through metrics of awareness of research findings on the part of participants in specific policy deliberations, as well as through media exposure. A monitoring and evaluation framework of this type requires early establishment and regular tracking. Important contributions of research efforts are likely to elude this approach and should be collected periodically as case studies. Contribution analysis is relatively new in the field of evaluation, but is increasingly being recognized as appropriate for assessing the impact of policy analysis (Riley et al. 2018). Contribution analysis can be as demanding as the underlying policy research even in developed countries (see Riley et al. for an example in the field of public health in Canada). The resource intensity of such an effort would not be appropriate for a low or middle-income country and would probably not be embraced by government clients or partners in the agricultural sector. However, modified or simplified approaches can be feasible and useful. Piloting of such approaches under an initiative to strengthen the flow of evidence into policymaking would be a methodological contribution over and above resulting changes in policy. 6.10 u The costs of supporting trusted agricultural policy analysis are modest Finally, the full costs of supporting the use of trusted agricultural policy 89 analysis in decision-making are modest. Underinvestment in agricultural policy research occurs not because the tasks are inherently expensive, but rather because the benefits are not fully demonstrated and perceived. Reaping tangible returns from the substantial effort to date to improve agricultural data and policy analysis in African countries requires investing in the use of the product for the purpose intended. To address this, any newly formed effort to promote agricultural data collection, policy research, or policy analysis should be clearly linked to a process that ensures attention to issues of the highest priority to key policy actors. Specific provision should then be made for linking analytical output to the timely provision of the briefing materials required by decision-makers. 90 7 92 Conclusions Agricultural decision-makers in Africa face a tightening environment for accountability to their constituents. Competition for resources is high and rising, as are the complexities of the agricultural policy agenda and demands on decision-makers. Countries that are best able to support their decision making with evidence-based policy analysis will do better in the long run than those that succumb to conjecture, uncontested demands of special interests, or ideology. Informed policy debates will require improved institutional capacity of public agencies and civil society organizations to curate, interpret, consolidate, refine, and use evidence. African countries are not uniformly positioned to move on this agenda, but some are clearly ready to do so as suggested in the summary of cases. Although this analysis will not apply to agricultural decision-makers in all African countries, it was strongly validated by the workshop reported upon in Annex I of the present paper. The historical models discussed in this review focus largely on the supply side of policy advice. That is, they focus upon how to educate for, organize, generate, and position policy research and analysis to assure that competent, credible, and trusted input will be available in a timely manner to inform specific policy deliberations. This focus on the supply side is an understandable consequence of the origin and intent of the efforts when they were created. In the current period, the participants in the workshop stressed the demand side of the equation such as the importance of political will, or the degree to which a government actively seeks to embed evidence in the policy process. This shift in circumstances suggests that newly organized efforts should be governed by the demand-pull of the policy process and its key agents. Government entities demonstrating strong political will should be favored in the allocation of logistical and financial support to move ahead. A call for expressions of interest as implemented under the FIRST program could invite responding governments to lead a design effort featuring specific analytical products and their communication to key actors in the policy process. The data summarized in Table 1 showing a rise in the relative weight of domestic funding of agricultural public expenditures in Africa suggests that there should be a corresponding rise in cost sharing of analytical work. Shared 93 funding is consistent with strong local ownership of the evidence cycle. Part of the response to a call for expressions of interest should include a willingness to commit to share costs. The lively and informed discussion held during the May 1-2 workshop in Washington included a full range of policy analysts, but no representatives from governments. This was deliberate, as the purpose was to promote a learning event to consult African analysts and not to get into discussions of present government policies or desires. Both the workshop and this paper found that the appropriate remedies to build institutional capacity for evidence- based policy are country-specific. In countries where government interest in the topic is strong, a process of country-led deliberation can yield an increased understanding of the policy process, the importance of evidence, shared views on how best to produce and deliver evidence, and a commitment to use of evidence. 94 References Alston, J., G. Norton, and P. Pardey. 1995. Science under Scarcity: Principles and practice for agricultural research evaluation and priority setting. Ithaca, N.Y.: Cornell University Press. 618 pp. Anson, R. and T. Mogues. 2016. A Systematic Review of Cross-Country Data Initiatives on Agricultural Public Expenditures in Developing Countries. IFPRI Discussion Paper 01541, Washington, D.C., IFPRI. July Available at: https://papers.ssrn.com/sol3/papers. cfm?abstract_id=2813410 Anderson, K., ed. 2009. Distortions to Agricultural Incentives: A global perspective 1955-2007. Washington D.C.: World Bank and Palgrave Macmillan. Alliance for a Green Revolution in Africa (AGRA). 2018. Africa Agriculture Status Report: Catalyzing Government Capacity to Drive Agricultural Transformation (Issue 6). Nairobi, Kenya: Alliance for a Green Revolution in Africa (AGRA) Bates, Robert H. 1981. Markets and States in Tropical Africa: The political basis of agricultural policies. Berkeley, CA: The University of California Press. Beintema, N., P. Pardey, and J. Roseboom. 1998. Educating Agricultural Researchers: A review of the role of African universities. EPTD Discussion Paper No. 36. Washington, D.C.: IFPRI and ISNAR Birner, R. 2018. Mid-Term Implementation Review of FAO’s Monitoring and Analyzing Food and Agricultural Policies program Phase II (MAFAP II): Final Report (unpublished). FAO, Rome. Brüntrup, Michael. 2011. The Comprehensive Africa Agriculture Development Program (CAADP)– An Assessment of a Pan-African Attempt to Revitalise Agriculture. Quarterly Journal of International Agriculture 50 (2011), No. 1: 79-106 CGIAR Independent Review Panel. 2008. Bringing Together the Best of Science and the Best of Development. Independent Review of the CGIAR System. Report to the Executive Council of the Consultative Group on International Agricultural Research. Washington, DC. Commander, Simon ed. 1989. Structural Adjustment and Agriculture: Theory and Practice in Africa and Latin America. Portsmouth, N.H.: Heinemann. Crawford, Eric; Duncan Boughton; James Allen; and John Staatz. 2009. Lessons learned from 25 years of policy research and outreach. Michigan State University International Development Working Paper 101. East Lansing, MI: Michigan State University. Delgado, Christopher and Sidi Jammeh, eds. 1991. The Political Economy of Senegal under Structural Adjustment (New York: Praeger Publishers). Delgado, Christopher. 1998. Africa’s changing agricultural development strategies: past and present paradigms as a guide to the future. The Brown Journal of World Affairs. Vol. V issue 1, Winter/Spring 95 Delgado, Christopher et al. 2011. Food Security: The Need for Multilateral Action. in S. Fardoust, Y. Kim and C. Sepulveda, eds. Postcrisis Growth and Development: A Development Agenda for the G20. (Washington. D.C.: The World Bank, 2011) 383-425 DFID (UK Department for International Development). 2013. Project Completion Review of Ethiopian Strategic Support Program (ESSP-II). November. Diribia, G. and C. Man. 2019. Building a Big Tent for Agricultural Transformation in Ethiopia. A Report of the CSIS Global Food Security Project. Center for Strategic and International Studies. Washington, D.C., April. Available at: https://csis-prod.s3.amazonaws.com/ s3fs-public/publication/190423_Ethiopia_WEB.pdf Eicher, Carl K. 2003. Flashback: Fifty Years of Donor Aid to African Agriculture. Revised version of a paper presented at an International Policy Conference “Successes in African Agriculture: Building for the Future”, sponsored by InWent, IFPRI, NEPAD and CTA, Pretoria, South Africa, December 1-3, 2003. Available at: https://rmportal.net/framelib/ donor-aid-to-african-agriculture.pdf Eicher, Carl K. 2006. The evolution of agricultural education and training: global insights of relevance for Africa. Staff paper 2006-26. Department of Agricultural Economics, Michigan State University, East Lansing, Michigan. Eicher, Carl K. 2009. Building African scientific capacity in food and agriculture. Review of Business and Economics 54(3):238-257. Gill, Indermit and Kenan Karakulah. 2018. “Sounding the alarm on Africa’s debt,” Brookings Institution, Washington, D.C., April 6, 2018. https://www.brookings.edu/ wp-content/uploads/2018/04/global_figure-1_-debt-in-sub-saharan-africa.png Goldsmith, Arthur. 1990. Building agricultural research institutions: transferring the Land Grant model to India and Nigeria. Boulder, CO: Westview Press. Goyal, Aparajita and John Nash. 2017. Reaping Richer Returns : Public Spending Priorities for African Agriculture Productivity Growth. Africa Development Forum;. Washington, DC: World Bank. © World Bank. https://openknowledge.worldbank.org/ handle/10986/25996 License: CC BY 3.0 IGO. Haggblade, Steven. 2013. Unscrambling Africa: regional requirements for achieving food security. Development Policy Review 31(2):149-176. Hazell, Peter, Frank Place, and Eric Tollens. 2018. “Taking Stock of IFPRI’s Experience with Country Programs,” Independent impact Assessment Report No. 45, December 2018, International Food Policy Research Institute, Washington, DC. Idachaba, F.S. 1997. Instability of national agricultural research systems in Sub-Saharan Africa: Lessons from Nigeria. The Hague: ISNAR. Isinika, A.C. et.al.. 2016. Agricultural Policy Analysis Studies in Tanzania: A Historical and Thematic Perspective with Implications on Future Policy Research for Crop Production and Marketing, Tanzania Journal of Agricultural Sciences Vol. 15 No. 1, 51-68. Jayne, T, J. Govereh, A. Mwanaumo, J. Nyoro, and A Chapoto. 2002. False Promise or False Premise? The Experience of Food and Input Market Reform in Eastern and Southern Africa. World Development. Vol. 30 (11), November, 1967-1985 Available at: https://doi. org/10.1016/S0305-750X(02)00115-8 96 Jayne, T., S. Babu, D. Boughton, S. Hendriks, E. Mkandawire, F. Meyer, J. Staatz, S. Liverpool-Tasie, E. Crawford, and K. Sawadogo. 2019. “Building Locally-Led Agricultural Policy Analysis Capacity: Lessons from Experience in Developing Countries”, Draft Manuscript, June 10, 2019. Kaneene et al. 2015. Sub-Saharan Africa’s agri-food system in transition. Journal of Agribusiness in Developing and Emerging Economies 5(2):94-101. Kherallah, Mylene; Christopher Delgado, Eleni Gabre-Madhin, Nicholas Minot and Michael Johnson. 2002. Reforming Agricultural Markets in Africa: Achievements and challenges. Baltimore, MD: Johns Hopkins University Press. Krueger, A.; M. Schiff: and A. Valdes. 1991. The Political Economy of Agricultural Pricing Policy, Volume 3 Africa and the Mediterranean. Baltimore, MD: Johns Hopkins University Press. Kuyvenhoven, Arie. 2014. “Impact Assessment of IFPRI’s Capacity-Strengthening Work, 1985-2010.” Independent Impact Assessment Report No. 38, December 2014. International Food Policy Research Institute, Washington, DC. Lele, Uma. 1975. The Design of Rural Development: Lessons from Africa. Baltimore, MD: Johns Hopkins University Press (paper). Lipton, M. 1977. Why Poor People Stay Poor: A study of urban bias in world development. Canberra: Temple Smith ; Australian National University Press Lowder, S. and B. Carisma. 2011. Financial resource flows to agriculture: A review of data on government spending, official development assistance and foreign direct investment. ESA Working paper No. 11-19 Food and Agriculture Organization of the United Nations (FAO), Rome, December. www.fao.org/economic/esa. Lütkenhorst, Wilfried and Horst Minte. 1979. The petrodollars and the world economy, Intereconomics, ISSN 0020-5346, Verlag Weltarchiv, Hamburg, Vol. 14, Iss. 2, 84-89 pp. Available at: http://dx.doi.org/10.1007/BF02930203 McGann, James G.. 2018. “2017 Global Go To Think Tank Index Report”. TTCSP Global Go To Think Tank Index Reports. 13. https://repository.upenn.edu/think_tanks/13. Mellor, John; Christopher Delgado; and Malcolm Blackie, eds. 1987. Accelerating Food Production in Sub-Saharan Africa (Baltimore: Johns Hopkins University Press). Mellor, J. 1998. “Foreign Aid and Agriculture-Led Development”, Chapter 3 in C. Eicher and J. Staatz, eds., International Agricultural Development, 3rd edition. Baltimore, MD: The Johns Hopkins University Press. Meyer, Ferdi. 2018. “Agricultural Economists and Policy- Making: Agricultural Land, Policy and Scenarios,” Presentation at the 56th Agricultural Economics Association of South Africa Conference, Western Cape, South Africa, 25-27 September, 2018. Mink, Stephen. 2016. Findings across agricultural public expenditure reviews in African countries. IFPRI Discussion Paper 01522. IFPRI, Washington, DC. April. Ndulu, Benno and Nicolas van de Walle, eds. Agenda for Africa’s Economic Renewal. New Brunswick, N.J./Oxford: Transaction Publishers, 1996. ix + 246 pp. 97 Onyekwena, C. and Mma Amara Ekeruchwe. 2019. “Is a debt crisis looming in Africa?” Brookings Institute, Africa in Focus. Organization of African Unity. 1980. Lagos Plan of Action for the Economic Development of Africa 1980-2000. OAU, Addis Ababa. April. Available at: https://www.merit.unu.edu/ wp-content/uploads/2015/01/Lagos-plan-of-action.pdf Özgediz,S.. 2012. THE CGIAR at 40: Institutional Evolution of the World’s Premier Agricultural Research Network. Montpellier, France and Washington, D.C.: Consultative Group on International Agricultural Research, 2012. Paddock, W. and P. Paddock. 1967. Famine 1975! America’s Decision: Who Will Survive? New York: Little Brown and Co. Peters, M. 2009. “Agricultural Commodity Price Spikes in the 1970s and 1990s: Valuable Lessons for Today”. Amber Waves Magazine. U.S. Department of Agriculture, Economic Research Service. Washington, D.C.. March Place, Frank and Peter Hazell. 2018. IFPRI Country Programs: Lessons from case study successes. IFPRI Discussion Paper 01739. July. Renkow, M., and R. Slade. 2013. An Assessment of IFPRI’s Work in Ethiopia 1995−2010: Ideology, Influence, and Idiosyncrasy. Independent Impact Assessment Report No. 36. Washington, DC: International Food Policy Research Institute. Resnick, D., S. Haggblade, S. Babu, S. L. Hendriks and D. Mather. 2018. “The Kaleidoscope Model of Policy Change: Applications to Food Security Policy in Zambia.” World Development 109 (September): 101−120. Riley, B., A. Kernoghan, L. Stockton, S. Montague, J. Yessis, and C. Willis. 2018. “Using contribution analysis to evaluate the impacts of research on policy: Getting to ‘good enough’”. Research Evaluation, Volume 27, Issue 1, January 2018, Pages 16–27, https:// doi.org/10.1093/reseval/rvx037. Sahn, David; Paul Dorosh, and Stephen Younger. 1997. Structural Adjustment Reconsidered: Economic Policy and Poverty in Africa. Cambridge, UK: Cambridge University Press. Sakane, Nome; Christopher Delgado, and Ed Rege. 2019. Workshop Report: Consultation Meeting on Strengthening African Agriculture Policy Capacity, Hosted by the World Bank Group (WBG) Agriculture Global Practice in Collaboration with the Bill and Melinda Gates Foundation (BMGF), Washington, D.C., USA May 1-2, 2019 (manuscript available on demand). Sands, F. B. 2008. Assessment of the Impact of the Ghana Strategic Support Program (GSSP): A Final Draft Report. Washington, D.C.: IFPRI. Staatz, J. and C. Eicher. 1998. “Agricultural Development Ideas in Historical Perspective”, Chapter 1 in C. Eicher and J. Staatz, eds., International Agricultural Development, 3rd edition. Baltimore, MD: The Johns Hopkins University Press. Todaro, M. 1977. Economic Development in the Third World: An introduction to problems and policies in a global perspective. London and New York: Longman. Traub, N. L. 2012. Bureau for Food and Agricultural Policy (BFAP): Your partner in decision making. The Futures of Agriculture. Brief No. 10 - English. Rome: Global Forum on Agricultural Research (GFAR). 98 United Nations Food and Agriculture Organization, Monitoring and Analysing Food and Agricultural Policies (MAFAP) Program (MAFAP). 2017. Revue des Dépenses Publiques Agricoles (AgPER Léger). Bamako, Mali. Unpublished presentation. United Nations Conference on Trade and Development. 1999. Foreign Direct Investment in Africa: Performance and Potential. New York and Geneva: United Nations UNCTAD/ ITE/IIT/Misc. 15 United States Agency for International Development. 2017. Mid-Term Performance Evaluation of the Feed the Future Innovation Lab for Food Security Policy. Washington, D.C., USAID, Feed the Future, October 14. 191 pp. Von Braun, Joachim; Tesfaye Teklu; and Patrick Webb. 1999. Famine in Africa: Causes, Responses, and Prevention Baltimore, MD: Johns Hopkins University Press, 1999World Bank.1981. Accelerated Development in Sub-Saharan Africa: An agenda for action. Washington, D.C., World Bank. Available at: http://documents.worldbank.org/curated/ en/702471468768312009/pdf/multi-page.pdf World Bank. 1986. Financing Adjustment with Growth in Sub-Saharan Africa, 1986-90. Washington, D.C.: World Bank Publication 6082. Available at: http://documents. worldbank.org/curated/en/330661468742793711/pdf/multi-page.pdf World Bank. 1991. Report and Recommendation of the President of the International Development Association to the Executive Directors on a Proposed Development Credit of SDR 60 Million to Burkina Faso for a Structural Adjustment Program (SAL I). Washington, D.C., World Bank. Available at: http://documents.worldbank.org/ curated/en/324831468016863353/pdf/multi0page.pdf World Bank. 2001. Implementation Completion Report (IDA-35370) on an IDA credit in the amount of US$24.5 Million to The Republic of Tanzania for an Agricultural Sector Management Project-ASMP (Credit No. 2537-TA). December 31. Report No. 23354. World Bank. 2005. “Improving Public Sector Governance: The Grand Challenge?”, Chapter 9 in Economic Growth in the 1990s Learning from a Decade of Reform. Washington, D.C., World Bank. World Bank. 2007a. Project Performance Assessment Report and Sector overview United Republic of Tanzania Agricultural Sector management project (Credit 2537), Second Agricultural Research Project (Credit 3036), national Agricultural Extension Project II (Credit 2899). June 18. Report No. 40109-TZ. World Bank. 2007b. Agriculture for Development: World Development Report 2008. Washington, D.C.: World Bank.Zartman, I. W., and Christopher Delgado, eds. The Political Economy of Ivory Coast (New York: Praeger, 1984). 99 A 100 Annex I Executive Summary of Workshop Report May 23, 201930 Consultation Meeting on Strengthening African Agriculture Policy Capacity Hosted by the World Bank Group (WBG) Agriculture Global Practice in Collaboration with the Bill and Melinda Gates Foundation (BMGF) Washington, D.C., USA May 1-2, 2019 The Bill and Melinda Gates Foundation (BMGF), the Alliance for a Green Revolution in Africa (AGRA), and the World Bank Group (WBG)’s Agriculture Global Practice convened a consultation meeting (a learning event) on Strengthening African Agriculture Policy Capacity at the World Bank Group offices in Washington, D.C., USA, on May 1-2, 2019. The Consultation Meeting gathered some 43 representatives from key partners working on evidence- based policy support in the agriculture sector in Africa. Institutional categories represented included: African Union (AU) agencies (AU Commission and Arica Union Development Agency-NEPAD); international agencies and civil society organizations; agriculture policy analysis and education service providers active in Africa; and national economic and social policy institutes from Africa. The objective of the meeting was to get views of key stakeholders on ways to increase the use of hard evidence in agricultural policymaking in Africa. It sought to address four main areas: reflections on attempts made in recent decades to strengthen the capacity of governments to better incorporate evidence in agricultural policymaking and what can be learnt from them; what should be considered to constitute success by such delivery vehicles or models; the pre-conditions or underlying contexts that must be present to achieve success; and support or investment that would be needed to put in place working and sustainable interventions in this domain. In the 1980s and 1990s, donor partners helped a number of African countries to build agricultural policy analysis, monitoring, and evaluation systems. In some 101 cases, these donors made these policy support entities a precondition for donor funding of agriculture. Such funding at the time constituted a significant share (typically more than half) of agricultural public spending in most countries of the region. Since then, foreign funding of agriculture public expenditures in Africa has declined significantly and is typically at less than 15% of total. Unfortunately, starting in the late 1990s, most of the policy analysis/advisory units previously funded by donors disappeared once donor funding dried up. To meet overall national goals within a limited public budget that is spread over many government functions, African ministries responsible for finance are increasingly scrutinizing public funding and programming, including those of Ministries of Agriculture (MoAs), putting emphasis on expenditures that yield or have potential to yield greatest returns. In any event, both African ministries of finance and ministries of agriculture in an increasing number of countries are presently expressing interest in having better capacity to access and use evidence-based agricultural policy analysis from trusted sources in a format and institutional arrangement that serves national agricultural decision- makers in a timely fashion. The two-day meeting was designed as a highly participatory learning and sharing event, with Day 1 focusing on reflections based on experiences to date and Day 2 on an attempt to develop fit-for-purpose institutional arrangements to deliver the needed policy support going forward, based on lessons learnt. The workshop delivered two streams of output: a) lessons to date and implications; and b) principles that should guide design of institutional architecture for both current and future policy support. Lessons learned to date • Country context has a large role in determining needs and demand for use of evidence-based agriculture policy analysis and the responses observed. • Countries differ in capacity and differences are reflected in how countries generate and use evidence to inform policymaking. Low absolute capacity is not necessarily reflected in poor performance in policy analysis and use of evidence base. Such issues as innovative institutional arrangements driven by institutional leadership come into play. Capacity of overall agricultural policy systems has multiple dimensions including: the 102 availability and use of data and tools; human capacity; research, analysis; consolidation and delivery; decision-making processes; monitoring and evaluation of implementation; and internalization of lessons-learnt. It was considered that various agricultural policy system capacity investments over the past decades have without a doubt contributed to policy system capacities in some countries. However, having system capacity in place is not necessarily reflected in how it is used and the results of that use. • Capacities that are most constraining. Participants agreed that lack of capacity for use of evidence-based policy analysis in actual decision-making processes in agriculture greatly constrains the policy cycle. Capacity gaps exist on both the demand and the supply sides of using evidence for policy. From the standpoint of ministries, gaps on the demand side for evidence-based policy analysis include the lack of capacity to prioritize, the lack of effective brokerage and inter-sectoral coordination of approaches, poor leadership, and the lack of political support. On the supply side, they include limited availability of/access to critical data and tools, the inability to attract and retain necessary talent, the lack of connection to policy research organizations, limited capacity to generate and package evidence, poor communication and articulation of evidence, and low capacity of non-states actors to use evidence for advocacy. • Other key elements besides capacity include leadership and political will/commitment, having a critical mass of real champions for policy change, accountability and performance management systems, inter- sectoral coordination and quality engagement of key stakeholders through effective platforms that can serve to achieve alignment of development partners and the private sector with country/regional priorities, continuity and stability in key positions (especially Ministers and Permanent Secretaries), inclusivity mechanisms, and the ability to handle vested interests. Strong institutionalization of a policy support service/unit can help mitigate the impact of change of senior leadership at Minister and Permanent Secretary levels. • Ministries responsible for agriculture have a greater stake in and pay increased attention to using evidence-based policy analysis. The major ways in which the increased stake is being manifested include: the need to shore up the case for agriculture as a driver for economic growth and 103 increased need for hard evidence to present to the Ministry of Finance; an emerging accountability landscape including data, credibility, and honesty, continental level accountability mechanism—the Joint Sector Review; global drivers including climate change, increased demand for food, markets, and youth employment that require a more agile use of evidence; and decentralization and devolution of governments, changing complexity, and expectations from agriculture. • Sustainability drivers for a policy support service entity. These include having in place a functional performance management and accountability system; design and implementation that is relevant to local context; local ownership and engagement to ensure sustainability through short-, medium-, and long-term planning; and diversification of funding sources that support the entity. The unit can only be sustainable if it has annual budget allocation for its regular operations, with use of donor/project funds being limited to time-bound initiatives with specific deliverables that further contribute to sustainability or are intended to deliver a specific set of outputs. For sustainability and stability of the entity, it is desirable to align it more to the civil service than to transient political space. Finally, the service/entity should be able to coordinate presentation of evidence- based policy analysis among different stakeholders in the agricultural policy ecosystem. ‘Principles’ for institutional architecture going forward Participants worked in six breakout hypothetical country contexts groups: 1. “Nigeria” (a hypothetical large country representing a federal structure and significant mineral exports); 2. “Burkina Faso” (small, landlocked, largely semi-arid case); 3. “Tanzania” (medium sized country with strong agricultural growth interest); 4. “Kenya” (medium sized country with advanced agribusiness sectors as well as smallholders); 5. “Malawi” (densely populated small landlocked agricultural country with strong history of State intervention in agricultural markets); and 6. ‘Wakanda’ (a fictional country where participants were free to describe the context for agricultural policy as they wished). The assignment was for each group to come up with a policy support vehicle or delivery mechanism that is fit-for-purpose for the country context in question. Each country was considered to represent a typology of several countries that have similar contexts. Thus, the outputs were not necessarily 104 for the named case country, but rather for a typical category of countries. A country considering arrangements that could work for them could also adopt and adapt a combination of elements from the proposals made for multiple case countries. The key guiding principles that emerged from the group work generally characterize the policy support institutional architecture proposed for the six cases are: • Where to establish the unit. Strengthen existing institutions, with some tweaks to increase generalizability of lessons, wherever it applies. Refrain from creating new ones. Establish the unit to be closely linked to the Ministry of Agriculture and to be trusted by Ministry leadership for visibility purpose with some level of independence to avoid total capture by government bureaucracy. Noting the tendency to bow to political pressure to fund policy units located in the office of the President or Prime Minister, participants agreed that this should be avoided. The advantage of this approach includes close supervision by and resulting accountability to the highest political office. This may work well for the period the President or Prime Minister that established the unit remains in power. Several participants pointed out that when power changes hands, there is no guarantee that the new administration will sustain such a unit, let alone maintain the same priorities for which the unit was created. A better model is one in which the unit is strengthened and institutionalized to deliver on its mandate. That way the unit may still give due priority to supporting the vision of the current leadership without being wasted when political leadership changes. • Remit/mandate. Keep the entity/unit (the delivery vehicle/mechanism) broad enough to cover other ministries and departments that are effectively involved in the agriculture sector. It should not be designed to solely serve the ministry as an administrative entity, but agriculture as a sector/function to which multiple administrative departments and Ministries contribute. • Functions of the unit. The role of the delivery vehicle should be to curate, synthesize, and communicate evidence on policy options and their likely consequences in due form and in a timely manner to policymakers. Policy options should be informed by existing policy analysis processes to enhance evidence-based decision-making by policymakers. Importantly, 105 the unit may not necessarily need to have the capacity to generate the underlying policy analysis, but rather to connect to and rely on credible policy research institutions and think-tanks in the country. The unit should, however, be able to coordinate among the line ministries and implementing sectors across all levels. • Staffing of the unit. The unit/vehicle needs to be led by a senior and experienced professional with technical, managerial, and political skills. Other staff of the unit should be of high technical caliber to generate credible analysis to guide decision-makers. Being aware that such staff may eventually move to other organizations, it is important for the unit to have an on-going professional development component to ensure a pipeline of capacity enhancement for younger professionals. • Sustainability. In addition to building on existing institutions, create a government budget line or vote to meet the core expenses of the unit, including staff salaries and benefits. Donor financing can be sought for time-bound programs, if necessary. The required level of sustainability may also be attained on the long run in a stepwise process during which dependence on external (donor) funding is reduced until the unit is weaned off. • Measures of success. Short-term measures of success include the number and quality of policies adopted and implemented, and number of laws and regulations that improve the agriculture business environment. These should be tracked using a results framework, which needs to be developed at the time of establishment. Long-term success will include improvement in high level outcomes to which agriculture contributes including: jobs, farmer incomes, poverty reduction, and reduced hunger and malnutrition. Other success factors include the ability to coordinate the presentation of evidence-based policy analysis among different players in the agricultural policy space. 106 Endnotes 1. “Africa” is used as a synonym for “Sub-Saharan Africa” throughout this paper for simplicity and succinctness. 2. The assertions and figures given in this section summarize broader arguments in the full paper, which provides full citations on sources that are not repeated here. 3. See Alston, J., G. Norton, and P. Pardey. 1995. Science under Scarcity: Principles and practice for agricultural research evaluation and priority setting. Ithaca, N.Y.: Cornell University Press. 618 pp. 4. Jayne et al. (2019) point out the number of Ph.D. degrees in agricultural economics held by Africans working in Africa is thought to have expanded by a factor of 8 between 2004 and 2016. Examples of the sorts of evidence needed include, but are not limited to, comparisons of full costs and benefits of different agricultural policy options in terms of income and fiscal impact, panel-based assessments of who is affected and in which ways by policy changes and investments, impacts on resilience to shocks, and the efficiency of use of public expenditures in the sector. 5. The Kaleidoscope Model developed at IFPRI provides a state-of-the-art framework for considering these numerous factors, see Resnick et al. (2018). 6. An influential and sensationalist book at the time in this regard was Paddock and Paddock (1967). 7. This led to acrimonious debates at the time, often on ideological grounds, in addition to disagreement about what the relevant facts were. A lengthy exposition is found in Commander (1989), the multi-volume series Krueger, Schiff, and Valdes (1991) and an update in Anderson (2009). 8. The net taxation of agriculture to promote industrialization—if not the actual use of fiscal resources thus generated in Africa for services and construction—was very much aligned with development wisdom at the time (Todaro, 1977). 9. The Lagos Plan of Action was drafted in conjunction with an OAU meeting involving several heads of state in Lagos in April 1980 and largely reflected the views of Ministers of Planning, while the Berg Report was undertaken in response to a request by the African governors of the World Bank in 1979 (the African governors being either Ministers of Finance or Central Bank heads in their respective countries, the members of government responsible for financing the expenditures of other ministries). The intense and widespread policy interest in the differences of the mainly inward-looking view of the Lagos Plan of Action and the unabashedly outward-looking Berg Report at the time led to much of the academic and gray literature on these differences devolving into advocacy for one viewpoint or the other, rather than dispassionate and comparative analysis (Commander, 1989; Delgado, 1998). With the advantage of time and a large number of contributions, the separate online Wikipedia articles on both reports are presently among the most accurate and fair-minded presentations of the specific points (Wikipedia accessed June 15, 2019). 10. Senegal in 1979 was one of the first cases of an explicit Structural Adjustment loan for budget support from the World Bank. See Delgado and Jammeh (1991). 107 11. Anecdotally, a former World Bank Vice President for Africa, who had been a junior economist in the Senegalese MoEF during Structural Adjustment (and later become Minister of Finance), recalled on several occasions in large staff meetings at the World Bank the critical importance for national self-determination of having evidence based policy analysis from the MoAs available during negotiations with donors, and how much that improved after an APME type unit was established in the Senegalese MoA during the Structural Adjustment era. 12. Latin American and small Asian countries were also severely affected by the same global shocks impacting Africa in 1979, but countries like Mexico were able to formulate and implement stringent national strategies to get out of balance-of- payments difficulties promptly, while few African countries could or at least did the same (Commander, 1989; Krueger et al. 1991) 13. Examples that come to mind are Uganda (EPRC) and Kenya (KIPPRA); see Ndulu and van de Walle (1996). The Economic Policy Research Centre (EPRC) was founded in 1993 as an autonomous not-for-profit organization. As stated on its website, its purpose is “to fill fundamental voids in economics research, policy analysis, and capacity building for effective in-country contributions to Uganda’s policy processes” (See: https://www.eprcug.org/about-eprc.) The Kenya Institute for Public Policy Research and Analysis (KIPPRA) is an autonomous public institute initiated in May 1997 and eventually established under an Act of Parliament. Both EPRC and KIPPRA primarily serve their respective MoEFs. 14. As witnessed by the creation of a network to support 18 think tanks in eight African countries in 2004 through the ThinkTank Initiative in common with another 27 think tanks in Asia and Latin America (see: http://www.thinktankinitiative.org/think- tanks/sub-saharan-africa). 15. When one of the present authors was a U.S. Peace Corps Volunteer assigned to the Ministry of Planning in Chad in 1970, all of the technical departments of the Ministry of Planning were run by foreign embedded advisors with UNDP contracts. 16. There are a number of other prominent regional economic or agricultural policy analysis networks and training programs, including for example the Regional Universities Forum for Capacity Building in Agriculture (RUFORUM), the African Economic Research Consortium (AERC), and the new Alliance for African Partnerships. These are related to building capacity for policy analysis. However, they are not directly engaged in supporting the capacity of governments to use existing evidence in the same sense as the case examined and are not considered here. 17. World Bank Tanzania Credit No. 2537-TA 18. Crawford et al. (2009) is a general reference for this section. Additional material is found in Eicher (2006 and 2009), and Haggblade (2013). Interested readers are also referred to further discussion of the same institutions in Jayne at al. (2019). These authors were all architects (with others) of Michigan State University’s many contributions to agricultural policy education in Africa and--with the exception of the pioneer in this field, the late Professor Carl Eicher--are all still active today. 19. Case study based on publicly available project materials, unpublished reports, and stakeholder interviews by one of the authors. 108 20. Joint Sector Reviews (JSRs) are a well-established government-led process for different sectors in Africa originally initiated at the urging of external development agencies. They bring different stakeholders together to engage in dialogue, review status, and monitor expenditure, progress, and performance in the implementation of national sector plans or countries’ sector implementation frameworks. They are now institutionalized and owned in African agriculture through their inclusion in the CAADP framework and are a tool employed by ReSAKSS (see: http://conference. resakss.org/2018/07/25/strengthening-caadp-mutual-accountability-processes/). 21. Case study based upon independent Mid-Term Review of the MAFAP program (Birner (2018), not published), publicly available MAFAP materials, and stakeholder interviews by the authors. Also see Anson and Mogues (2016). 22. Case study based on publicly available FIRST materials and stakeholder interviews by the authors. See: http://www.fao.org/europeanunion/eu-projects/first/en/ 23. Case study based on publicly available TBI materials (see https://institute.global/) and stakeholder interviews by the authors. 24. Sources for this case include Diribia and Man (2019), an unpublished review for the Bill and Melinda Gates Foundation in 2012, and stakeholder interviews by the authors. 25. See http://www.agrodep.org. 26. See http://www.renapri.org/ for an overview of ReNAPRI’s organizational structure and activities. 27. This is the list as of June 2019 on the ReNAPRI website mentions only institutes in Eastern and Southern Africa plus Congo, but Jayne et al. (2019) mention the membership of the University of Ghana’s Institute for Social Science and Economic Research since 2018 and the possibility of future expansion to include other institutions in Nigeria and Ghana. 28. Through the AfDB representative to the workshop whose take-aways are summarized in Annex I. 29. See, for example, the call for proposals for an independent consultant to undertake a review of salaries and benefits at the ATA in 2017: https://ngojobsinafrica.com/job/ contractor-conduct-salary-benefits-survey-ethiopian-agricultural-transformation- agency 30. Report prepared by Nome Sakane (WBG), Christopher Delgado (WBG), and Ed Rege (ECI-Africa), cleared by Holger Kray, Head AAPU, WBG. 109 111 112